Re: [Asrg] VPNs vs consent

[Jose-Marcio Martins da Cruz <[e-mail]>]

Claudio Telmon wrote:
> Jose-Marcio Martins da Cruz wrote:

> 
> I need some more clarification on this, maybe it's just me not knowing
> enough about MTAs internals. As the border MTA receives a RCPT TO for
> either of these addresses, it should be able to know if it is a valid
> address. In my (very limited) experience, this means that each of these
> addresses mut be defined as a valid address, the MTA doesn't have rules
> to decide that since jose-marcio.martins_da_cruz is a valid address,
> jose-marcio.martins must be valid too. So, if each of these addresses is
> individually defined in some list/database accessed by the MTA, then
> with the same rules, the related token database should be accessed too.
> Should an "automatic aliasing" rule exist, then the same rule could
> exist for the token database. Also, if "somewhere" an alias is defined
> for an address, then the correspondent database could just be a pointer
> to the database of the "main" address. This could even implement "chains
> of aliases" as "chains of pointers to token databases".

The border MTA surely knows a list of valid addresses, but it may not know, all the time, 
that all this addresses resolve to the same login - sometimes some addresses are resolved 
in the final internal servers.

But well, you can find some organisations with this kind of thing nowadays. Don't know if 
this shall be taken into account to design future systems.

> I think this is feasible with an appropriate address book manager.
> Anyway, the load is for the MUA, not for the MTA, so the number of users
> shouldn't matter.

Hmmm. If the border MTA accept and the MUA reject by lack of consent, a bounce is generated.


-- 
  ---------------------------------------------------------------
  Jose Marcio MARTINS DA CRUZ           http://j-chkmail.ensmp.fr
  Ecole des Mines de Paris
  60, bd Saint Michel                      75272 - PARIS CEDEX 06
  mailto:[e-mail]