[geeks] UNIX development and makefile discussion

[microcode at zoho.com]

I'm writing some UNIX code that consists of a dozen or so source files for
the first time. I am ok with setting up makefiles for small projects of a
few files in one directory but things are starting to get out of hand. I
started making subdirectories to develop programs that are functionally
related to each other. This isn't working either because I'm finding I have
common code that's used by the various components and I haven't thought of a
good way to organize this.

I'd like to hear from people who are comfortable working on medium sized
UNIX applications about workflow. I was thinking about recursive makefiles
and I think I understand how to set them up but now I am considering that
given my components rely on common code that I have in another subdirectory
maybe this is not a good way to work.

How do you organize projects that consist of multiple main programs and
various functional pieces, some shared, some dedicated to a specific main
program? The code should work cross platform and should be able to build on
Solaris and Linux at least. Aside from writing portable code what should I
be considering?