Matt Saladna:
> Hello,
>
> When specifying a range of responses to ignore in postscreen_dnsbl_sites
> it appears that if a weight is zero it is ignored in favor of a non-zero
> weight.
Coming back to this thread, please ignore my previuous responses
about order dependence. They were wrong.
Sim
Matt Saladna:
> Is there any difference other than cognitive load between the two forms?
>
> postscreen_dnsbl_sites =
> zen.spamhaus.org=127.[0..255].[0..254].[0..255]*2
> zen.spamhaus.org=127.255.255.[252;254;255]*0
This explicitly assigns weights.
> versus
>
> post
/> The implemenation is order-dependent./
On 12.03.22 11:56, Wietse Venema wrote:
It does store the configuration in reverse order. However upon
closer reading of code that I haven't touched in 10+ years...
You are correct in that it applies all patterns that match. The
implementation simply a
On 12.03.22 11:50, Matt Saladna wrote:
Is there any difference other than cognitive load between the two forms?
postscreen_dnsbl_sites =
zen.spamhaus.org=127.[0..255].[0..254].[0..255]*2
zen.spamhaus.org=127.255.255.[252;254;255]*0
versus
postscreen_dnsbl_sites =
zen.
Matt Saladna:
> For Wieste,
That is WieTSe, if you don't mind.
>
> /> The implemenation is order-dependent./
It does store the configuration in reverse order. However upon
closer reading of code that I haven't touched in 10+ years...
You are correct in that it applies all patterns that match. T
On 2022-03-11 at 22:34:14 UTC-0500 (Fri, 11 Mar 2022 21:34:14 -0600)
Matt Saladna
is rumored to have said:
Spamhaus began flagging Cloudflare's servers, 1.0.0.1/1.1.1.1 as
public resolver resulting in the error message. Other DNSBLs pick up
responsibility, so the judgment shouldn't rely square
On 2022-03-11 at 17:20:41 UTC-0500 (Sat, 12 Mar 2022 09:20:41 +1100)
Phil Biggs
is rumored to have said:
Should the 127.255.255.[0..255] return codes really be weighted
zero, given that they indicate an error?
Absolutely.
With .254 being use of
a public/open resolver:
https://www.spamh
Title: Re: postscreen_dnsbl_sites precedence
Saturday, March 12, 2022, 2:37:15 AM, Matt Saladna wrote:
Hello,
When specifying a range of responses to ignore in postscreen_dnsbl_sites it appears that if a weight is zero it is ignored in favor of a non-zero weight.
mail_version=3.5.9
Matt Saladna:
> postscreen_dnsbl_sites=zen.spamhaus.org=127.255.255.[252;254;255]*0
> zen.spamhaus.org*2
The implemenation is order-dependent. Postscreen maintains a
list for zen.spamhaus.org, where the last entry appears first:
zen.spamhaus.org:
pattern=empty, weight=2
p
On Fri, Mar 11, 2022 at 09:37:15AM -0600, Matt Saladna wrote:
> When specifying a range of responses to ignore in postscreen_dnsbl_sites
> it appears that if a weight is zero it is ignored in favor of a non-zero
> weight.
No. Rather, when the same source is listed twice, the weights are
added,
10 matches
Mail list logo
