| 1 |
On Mon, May 11, 2015 at 04:47:31PM -0400, Michael Orlitzky wrote: |
| 2 |
> On 05/11/2015 04:08 PM, Robin H. Johnson wrote: |
| 3 |
> > By drop, I will clarify that they should ideally be rejected at SMTP |
| 4 |
> > time, not silently dropped. |
| 5 |
> |
| 6 |
> I believe those logs show a rejection after the message has been |
| 7 |
> accepted initially (if I'm wrong, you can ignore the rest of this). |
| 8 |
|
| 9 |
The analysis is correct. Pre-queue filtering will help as we can safely |
| 10 |
-meaning without causing backscatter- lower the threshold we reject spam |
| 11 |
at. There will still be some spam making its way to gmail but perhaps |
| 12 |
it will be low enough to stay under gmail's radar. |
| 13 |
|
| 14 |
The correct solution is to stop forwarding spam and the easiest way is |
| 15 |
just stopping forwarding. There are valid policy reasons for not going |
| 16 |
that route but continuing forwarding because it is too difficult to |
| 17 |
configure gmail is, well, not something I'd be comfortable with. I do |
| 18 |
expect more from gentoo devs. |
| 19 |
|
| 20 |
In this case (in most cases?), infra should not be looking for consensus |
| 21 |
but rather do what is right. |
| 22 |
|
| 23 |
Anyway, I believe infra has all the info it needs at this point and I am |
| 24 |
fine with whatever decision they make. |
| 25 |
|
| 26 |
-- |
| 27 |
Eray |
Archives