[webdev] cvs access

[Tom Hull <thull@xxxxxxxxxxx>]

OK, I have set up the CVS on the new webserver to provide anonymous
read access and read/write access using ssh. This is a critical step
in that it allows people to get their own copies of the openacs files
used on the web server, set up your own private copy of the web server,
hack on the files, and (if you have a regular account on the server)
check your files back in. This whole process could be made slicker,
but at least I have it working here.

If you have an account on the aclug web server (which AFAIK only
Dale, John A., and myself have), all you need to do to get the
current code is:

  $ mkdir ~/aclug
  $ cd ~/aclug
  $ export CVS_RSH=ssh
  $ # replace YOU below with your user account name on the server
  $ export CVSROOT=:ext:YOU@xxxxxxxxxxxxxxxxx:/home/cvs/cvs
  $ cvs checkout openacs

If you don't have an account, you can use this:

  $ mkdir ~/aclug
  $ cd ~/aclug
  $ export CVSROOT=:pserver:anonymous@xxxxxxxxxxxxxxxxx:/home/cvs/cvs
  $ cvs login
  # hit enter for empty password
  $ cvs checkout openacs

I started to try an approach using ssh for anonymous access, but thus
far haven't been able to get it working. The instructions are in:

  http://www.kitenet.net/programs/sshcvs/

To run your own copy of the server, you'll have to:

 1) Install and initialize postgresql.
 2) Install aolserver, including the postgresql module.
 3) Get a copy of the aolserver init file, and edit it to specify
    your own database, database password, and where your openacs
    files are. If you want to run another web server on the same
    machine, you'll have to edit the port numbers.
 4) Get the openacs files from our cvs (above).
 5) Load up the database schemae.
 6) Create some user accounts and dummy up some data.

Most of this is documented in my earlier email postings. Sooner or
later I'll document this better. (The more people who try to do
this, get baffled, and ask questions, the sooner. Natch.) Maybe
we'll concoct a standard test database; maybe a configurable program
to build test databases (e.g., if you want to test something with
1000 users). Meanwhile, if you need an aolserver init file, write
me and I'll mail you one back.

The old cribsheets that I wrote up are at:

http://www.complete.org/mailinglists/archives/webdev-200203/msg00038.html

http://www.complete.org/mailinglists/archives/webdev-200203/msg00039.html

The first covers server configuration; the second (much more schematic)
covers openacs modules.

-- 
/*
 *  Tom Hull * thull at kscable.com * http://www.tomhull.com/
 */