[PLUG] please stop modifying the message body

[Jason Barbier]

On Fri, May 28, 2021, at 3:54 PM, Tom wrote:
> On Wed, 26 May 2021 15:08:15 +0000
> Ben Koenig <techkoenig at protonmail.com> wrote:
> 
> > I went through and compared headers from a bunch of different senders
> > to the list. they all list DKIM as failed, but only yours (Tom's) get
> > marked as spam routinely. I started looking up source IPs on
> > mxtoolbox.com, and nuegia's (23.92.27.105) is listed, while others
> > aren't. this might be a reason for the difference. unfortunately this
> > particular blacklist is not one we can just request removal from,
> > it's at the ISP level. -wes
> > 
> > That matches what gmail is saying about the messages. The TLS
> > handshake succeeds, but a message is added from google claiming
> > "google could not verify that it actually came from nuegia.net"
> > 
> > When I click "learn more" it sends me to a generic help page.
> > 
> > Tom, can you send another reply to this thread? I wanna see how
> > protonmail handles it.
> Sure thing, but before we blame Gmail and as much as Google Corporation
> is a tumor on the global email system I don't think it's right to place
> all the blame there at the moment.
> 
> If DKIM is indeed failing, (probably due to message body tampering) I
> believe Gmail is in the right here to say that they cannot verify my
> mail came from nuegia.net.
> 

Except it is, and its a well known issue, https://wiki.debian.org/OpenDKIM, https://wiki.list.org/DEV/DKIM
A lot of lists get around it by either a.) stripping all dkim from all messages b.) ignoring the problem c.) rewrite the from address and resign the message. DKIM is flawed in its basic design when it comes to external mail lists and no amount of not rewriting your message fixes it sadly because then the next thing that breaks your dkim signatures after the body rewrite is stopped is the header rewrite to add reply-to and the mail list headers which have to be there for a bunch of reasons some of which are legal.

> I would greatly appreciate it if the operators of this list could at
> least test trying to make DKIM work. From what I read of the specs, the
> 'Sender' header, and my experience with using other mailing lists I do
> not believe DKIM to be incompatible with mailing lists.
> 
> That being said this UCEPROTECT-L3 company is a huge scam causing
> collateral damage and Google has their spam rule anomaly scores all
> screwed up.
> 
> Not entirely ruling out the possibility that Google anti-competitive
> actions are a part of why this is happening, as I do have an article up
> that criticizes Google Chrome and the WhatWG. Just not claiming it's
> the whole reason.
> 
> -- 
>  _______________________________________ 
> /  Against his wishes, a math teacher's \
> | classroom was remodeled. Ever since,  |
> | he's been talking about the good old  |
> | dais. His students planted a small    |
> | orchard in his honor; the trees all   |
> \ have square roots.                    /
>  --------------------------------------- 
> \
>  \
>    /\   /\   
>   //\\_//\\     ____
>   \_     _/    /   /
>    / * * \    /^^^]
>    \_\O/_/    [   ]
>     /   \_    [   /
>     \     \_  /  /
>      [ [ /  \/ _/
>     _[ [ \  /_/
>   
> 


---
GPG: https://corrupted.io/kusuriya.pub