automake
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Building prog first


From: Ralf Wildenhues
Subject: Re: Building prog first
Date: Sat, 27 Mar 2010 07:50:00 +0100
User-agent: Mutt/1.5.20 (2009-10-28)

A late hello,

* Robert J. Hansen wrote on Tue, Mar 23, 2010 at 12:12:46AM CET:
> On 3/22/10 6:50 PM, John Calcote wrote:
> > Reuben, you've just hit upon one of the two most significant problems
> > with Javadoc and the like (including doxygen, man pages, and info pages):
> 
> Agreed -- which is why I think it would be wonderful if there was strong
> Autotools support for literate programming.  Unfortunately, my m4-fu is
> not strong enough to figure out the appropriate macros to discover
> ctangle, cweave, etc.

One of the main issues with ctangle, cweave etc. is that they tend to
have an unpredictable set of output files.  Even just haveing more than
one output file by default is a bit difficult to map to 'make' semantics
(bug it's possible; see 'info Automake "Multiple Outputs"), but only
knowing the set of output files from looking at the *contents* of the
input files, and possibly the command line options of the rule to
generate the output files, works very much against 'make' and Automake.
I wish Knuth had considered that part a bit more.

That said, if somebody can come up with a decent scheme for how to
process things, that would be idiomatic for typical users of these
file types, I'm open to including rules for these in Automake.  If you
don't know how to translate this into m4 macros or a patch against
automake.in, decribe things in a way that we can understand what exactly
automake is supposed to do.  Typically, that already helps a lot for the
actual implementation.

Cheers,
Ralf




reply via email to

[Prev in Thread] Current Thread [Next in Thread]