Hello people,  

This report that you're about tor read has been written and was supposed 
to be mailed during the previous weekened but before doing so I wanted 
to get the 'OK' from my mentor which was too busy for the last few days.
Yesterday, we had a talk and since he seemed to be ok with what I had in
mind, I decided to mail it here as well. 

This week, I must admit that I’ve not done as much work on the project
as I could have, mostly because I am already ahead of schedule and also
because I’ve been thinking about how to develop the project from now on.

The project started as an implementation of OpenPAM modules for Linux
but what I’ve been aiming for was that the modules would be usable under
BSD systems as well, which would mean that there would be a single
package of pam modules for both systems. As the project progressed, I
found out that the BSD compatibility was not as simple as I’ve
originally thought it to be and that large chunks of code would have to
be rewritten, which would only bring more complexity to the project
(which comes in contrast with OpenPAM and its philosophy). In addition,
I’ve been thinking why would any BSD system would switch from the
current set of - lightweight- pam modules that are in use to another set
which wouldn’t be designed explicitly for BSD and would probably not be
as flexible as the original ones.

With all that in mind, I decided to restructure all the project and
remove any BSD-only code. From now on, the project will focus on
providing a working set of pam modules build with OpenPAM for all Linux
distros and on the side, I’ll try to write any pam module that the BSD
systems don’t already have from scratch and only for that target
OS.That’s pretty much what I’ve decided but, given that I haven’t
discussed it with my mentor yet, the decision may change (although not
much).On the bright side, I’ve completed for the project the following
modules :

* pam_rootok
* pam_nologin
* pam_securetty
* pam_deny
* pam_permit
* pam_unix (*)

* pam_unix’s base code has been finished but there are still some 
matters that need to be resolved.

Regarding the project timeline (since we are coming close to the
midterm evaluations), I amahead of schedule regarding the modules
produced but I really should debug and finish pam_unix as well as
soon as possible so that the modules produced can be tested as a 
set. Apart from the debugging process, documentation needs to be
written for each of the produced modules which will be heavily based
on the LinuxPAM and freebsd-lib/libpam documentation since all these
modules are similar to a great extend.

Regards, 
Seraphim
-- 
gentoo-soc@g.o mailing list