| 1 |
On Tuesday 20 April 2010 17:51:12 Harry Putnam wrote: |
| 2 |
> Grant Edwards <grant.b.edwards@×××××.com> writes: |
| 3 |
> > On 2010-04-20, Harry Putnam <reader@×××××××.com> wrote: |
| 4 |
> >> About all the snipes concerning hacking sendmail.cf... I'm sure you |
| 5 |
> >> are all aware that any hacking needs to happen in sendmail.mc... then |
| 6 |
> >> let m4 sort out sendmail.cf. |
| 7 |
> > |
| 8 |
> > IOW, sendmail has a configuration file so incomprehensible that the |
| 9 |
> > configuration file needs a configuration file. |
| 10 |
> |
| 11 |
> Internet mail is quite complex, yes. |
| 12 |
|
| 13 |
This statement is the source of the confusion surrounding sendmail. |
| 14 |
|
| 15 |
Internet mail is not complex, it is stunningly simple: |
| 16 |
|
| 17 |
mail comes in, |
| 18 |
look up where it should go, |
| 19 |
send it there |
| 20 |
|
| 21 |
In between you might hand the message off to virus and spam scanners, you |
| 22 |
might look up an ACL before accepting it coming in, but those are all |
| 23 |
additives to find valid mail. Remove the additives, and you get the amazingly |
| 24 |
simple lookup table scheme described above. |
| 25 |
|
| 26 |
There isn't even an inherent difference between relays and final destination |
| 27 |
MTAs, they still send the mail somewhere. The difference is in the viewpoint |
| 28 |
of the sysadmin. |
| 29 |
|
| 30 |
|
| 31 |
-- |
| 32 |
alan dot mckinnon at gmail dot com |
Archives