Aha!

That makes a lot more sense.  It's definitely a dumb way to do it.

I guess maybe they assumed everyone sending mail has at least a /24 all to 
themselves.

In our case our most-affected customers are in fact only on small portions of 
/24s that are shared with other users we don't have control over 
[facility-provided IPs].

I do understand that "ip neighbors" can cause issues in some cases but 
generally that's a blacklisting that is visible and can be disputed/resolved.

I wonder what the issue is this system was supposed to solve.  I could 
speculate but it's not helpful.

And just in case anyone missed it - I am sorry for coming across like a d*ck.  
I shouldn't have let my frustration get the better of me.

Thank you,

Michael Denney
MDDHosting LLC
http://www.mddhosting.com/

> On Feb 20, 2026, at 8:23 PM, Philip Paeps <[email protected]> wrote:
> 
> On 2026-02-21 07:49:10 (+0800), Michael Denney via mailop wrote:
>> I checked mail logs going back to December 2025 - and I see exactly zero 421 
>> errors.  As a matter of fact we have exactly zero 4XX responses from Outlook 
>> - network wide, but I do see 550's with the error 3150.
>> 
>> Are you sure this is properly implemented?
>> 
>> Feels half-baked as though it was AI vibe coded.
> 
> I suspect it's implemented on a network boundary that isn't a sender 
> boundary.  One sender in a /24 is getting the 4xx tempfails and an unrelated 
> sender in the same /24 gets a 5xx on first interaction.
> 
> That's clearly stupid but it's obvious that spam filters are no longer 
> maintained by sentient humans.
> 
> The good (?) news is that there are fewer and fewer humans who care about 
> email who host theirs at Microsoft.
> 
> Philip

_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop

Reply via email to