Re: testing email deliverability... [message #186047 is a reply to message #186037] |
Thu, 05 June 2014 10:59 |
AvdB
Messages: 2 Registered: July 2013
Karma:
|
Junior Member |
|
|
"The Natural Philosopher" <tnp(at)invalid(dot)invalid> schreef in bericht
news:lmoeok$5v6$1(at)news(dot)albasani(dot)net...
> what I want is a function like
>
> is_valid_mail("user12345(at)gmail(dot)com")
>
> that will return true or false if gmail accepts the user as known (yes the
> spam is always @gmail.com) WITHOUT having to write a dns mx lookup, then
> connect to socket 25 and etc etc etc..
>
> exim -bt tells me if the domain is valid, but not the user.
>
> I am not particular how its achieved - anything can be merged into the
> existing code.
As an idea to work on:
is_valid_mail() checks your database for known users.
If unknown, display a 'verify first' page.
Present a unique code which the user has to mail back to you.
Still not 100% save, but if a user can send using a certain address, there's
reason to believe he can receive as well. Add the user to your database.
Not enough?
Make it a three way handshake.
The user mails you his code, your autoresponder sends (after verifying the
code, obviously) another unique code.
Your form checks { is_valid_mail() or presents_valid_code() }
You now have proof email traffic is possible in both directions.
|
|
|