The Arusha Project is organized around `teams', and you will need to get different teams' packages/bits from different places. The core teams' packages come from SourceForge.

Arusha bits from SourceForge come by CVS or through the SourceForge `releases' mechanism. CVS is more the `Arusha style' (i.e. we recommend it), and hard-core developers will definitely want that.

The teams with bits from SourceForge include: ARK (core "mechanism" team), sidai (core developers' "policy" team), sample1 (Sidai-style test-drive team), `arksf1' (the team concerned with the ARK content at SourceForge), `verilab2' and `glasli2' (real `site' teams; SourceForge downloads only).

CVS as a user/curious-person: (read-only access)

Use the following commands to check out a team's CVS module:

# environment:
CVSROOT=:pserver:anonymous@cvs.ark.sourceforge.net:/cvsroot/ark
#
# commands:
cvs     login
# (and just hit "Enter" for the password.
#  You'll only have to do this once)
cvs -z3 checkout -P <team-name>

CVS as a developer: (write access)

# environment:
CVS_RSH=ssh     # NB: we hope you have SSH...
CVSROOT=<your SourceForge username>@cvs.ark.sourceforge.net:/cvsroot/ark
#
# commands:
cvs -z3 checkout -P <team-name>
# (and supply your SourceForge password when prompted;
# you'll be prompted at pretty much every cvs command)
# (note - no login command)  Examples:
# cvs -z3 checkout -P ARK
# cvs -z3 checkout -P sidai
# cvs -z3 checkout -P sample1

Downloads from SourceForge:

We supply various sets of bits via the SourceForge downloads mechanism: at our downloads page (or the prettier file releases page).

We put out releases (version numbers of the form n.m or n.m.o) or snapshots (version numbers are dates). The difference is not that important.

We have not done any `releases' as of July, 2001.

`Snapshots' (we have done many of these) tend to be sane. They are always announced on the mailing list(s), so you will know what's what if you are subscribed.

The downloadable files are named ark-<team-name>-<yyyymmdd>.tar.gz. Don't mix-and-match the dates. You can unpack a `set' to get a complete ARK tree:

for i in ark-<team-name>-<yyyymmdd>.tar.gz ; do
    gzip -dc < $i | tar xf -
done