It seem easy, but that prevented (at least in my case) to make the
system better.
The way we use multiple module in passdb is subtly broken and exposes us
to inconsistency and a lot of races.
and is not nice to have races in the users database.

the sam initially made a sane route and we also discusse dto not permit
anymore multiple backends at the same time, consistency and races where
my concern, but seem that thesse arguments have not stick

I still think sam is in the wrong direction, as multiple domain will
never be supported in samba so going that direction by default instead
of relegating the possibility in an module is wrong imho.

Plus I recently found another big problem in the interfaces that have to
do with race conditions, but that's another story ...

Simo.

On Fri, 2002-09-27 at 12:51, Andrew Bartlett wrote:
> Simo Sorce wrote:
> 
> > And in my honest opinion we should get out the possibility to have
> > multiple backends active at the same time, I really think that  move has
> > put us back 6 months in development and has caused more problems than
> > what the pros get with such a monster.
> 
> I still don't see where you draw such a conclusion.  In the pdb code,
> the multiple backends case just fell out of the design - it cost us very
> little indeed.  The same applies to the new SAM stuff, it didn't impose
> a significant design penelty, but was catered for.  (In the case of the
> SAM, each domain must have only one backend however).
> 
> Andrew Bartlett
> 
> -- 
> Andrew Bartlett                                 [EMAIL PROTECTED]
> Manager, Authentication Subsystems, Samba Team  [EMAIL PROTECTED]
> Student Network Administrator, Hawker College   [EMAIL PROTECTED]
> http://samba.org     http://build.samba.org     http://hawkerc.net
-- 
Simo Sorce - [EMAIL PROTECTED]
Xsec s.r.l.
via Durando 10 Ed. G - 20158 - Milano
tel. +39 02 2399 7130 - fax: +39 02 700 442 399

Attachment: signature.asc
Description: This is a digitally signed message part

Reply via email to