[gentoo-soc] BSD port of Gentoo (Portage tools)

[Patrice Clement]

Hi there,

I'm pasting the same mail I've sent to Davide Italiano, the
proposed mentor for the BSD port of Gentoo GSoC project [1].

"I write you this email to get futher informations about this GSoC
2009 project idea posted on the Gentoo wiki.

I'm interesting about the "porting Gentoo on BSD OSes" project;
recently, I was looking for a way to use Portage tools on FreeBSD (I
daily use FreeBSD on my desktop and Gentoo on my laptop), .. and what a surprise to see
this project on the Gentoo GSoC projects list.

I have read on the Gentoo/nbsd page that some work has been already
done. I have also read that you are one of the maintener of Gentoo/fbsd.
So here is my questions:
- On what BSD OS should the project has to be aimed ? Every *BSD OSes
or one in particular ?
- Porting Gentoo means porting the entire Portage tools (emerge,
equery, ... etc), is that correct ?
- The deliverables for this project are.. an entire modified OS with
 the tools ? the patches ? a LiveCD ?

I have other questions, of course, but with theses, I think I should
have a global approach on what are the needs of this project."

After some mail exchanges with Davide, I have submitted the idea of
coding a tool in a POSIX-compliant way (something using Bourne shell)
in order to do the same job as emerge does. Using Bourne shell means it
should be portable and so, painless to bring to a new platform.
I can read here [2] some work has already been done. I've tryed to see
the work achieved but it seems a bit wide (while downloading a huge stage
which weights around 100 Mb! [3]).

When targeting to port Portage a BSD, or whatever, platform, how do
you plan to manage that ? As I far as I can imagine your work, you copy
the tools (emerge, ebuild.sh, etc..) on the target platform and try to fix the
broken stuff. Of course, I don't think that it sounds easier as it
seems.

I really don't know where I'm going because I'm just a Gentoo user,
not an expert, that's why I want as much informations as I can have
in order to know exactly what should be done for such a project, and to
draft a sane proposal.

Another question: as some work has already been done for differents BSD
platforms (OpenBSD, NetBSD), I would like to know if I can plan to
start from the past work or if I have to start something new, from scratch ?
Second solution seems much more pretty, and if so, I would like to define in what
programming language it should be achieve (as I said, Bourne shell seems
to be a good candidate for the job).

I put the gentoo-soc in cc, perhaps some people may be able to provide me futher informations ? :)

Thanks in advance.

Patrice

[1] http://en.gentoo-wiki.com/wiki/Google_Summer_of_Code_2009_ideas#BSD_ports_of_Gentoo:_OpenBSD.2C_NetBSD.2C_DragonFlyBSD.2C_etc.
[2] http://www.gentoo.org/proj/en/gentoo-alt/bsd/index.xml
[3] http://dev.gentoo.org/~the_paya/7.1/