Re: Why SRS really sucks
"Dick St.Peters" <[email protected]> Tue, 28 Mar 2006 16:40:07 -0500
| Newsgroups | gmane.mail.spam.srs.general |
|---|---|
| Message-ID | <[email protected]> |
Tom Lahti writes: > There is no way in heck you should be pulling apart someone else's > rewritten local part and trying to make sense of it, and worse -- > doing something with "information" based on your assumptions that you > know what it means. Tom, I think you're forgetting that SRS is designed to have the LHS pulled apart in multi-stage forwarding, so the LHS doesn't grow out of bounds. A first rewriting results in an SRS0 address, while forwarding of that results in an SRS1 address. Forwarding of the SRS1 address results in another SRS1 address. To keep the address from growing too big, previous forwarders are eliminated from the encoded return path, and that requires forwarders to understand and modify the local part from previous forwarders. To be compliant, your SRS must enable this, and that requires it stick to the documented syntax. That in turn means your syntax can be interpreted by others for reasons of their own. If you try to prevent this by altering your syntax, then your mail won't pass through multi-stage forwarding with SRS. -- Dick St.Peters, [email protected]