Using ports tree as developement of new source

I am interested in using the ports environment as an inhouse build and
distribution system that will piggy back off the great freebsd port tools
like portinstall etc. This source and resulting port will be of no use
to others so will never be merged into the existing FreeBSD ports tree. I
have been reading the ports guide and it seems the ports environment revolves
around the use of existing tar balls and the fact one needs to 'fetch'
the package etc. This suggests that development is done elsewhere
and then massaged into the ports environment when development is complete.

What I would like is a separate area, either in the existing ports tree or
completely separate, where I can have a set of directories that house my
source code and I am able to develop and then package this source into
a port. I would then like to be able to install these ports onto remote
machines by setting the PKG_SITES and doing a portinstall -PPR on the
remote machine. Am I able to use the ports environment for what I want.
The first set of problems I found is the dependency to require a fetch.
Then the need to find and then unpack the tarball from distfiles.
 
Port system is made of a single Makefile written in BSD format that is located in /usr/ports/mk/bsd.port.mk other 3rd party routines are also present there.

You can reprogram it to use other tools that do same job. (fetch etc)
 
Back
Top