On 2021-09-22, Stefan Claas <spam.trap.usenet@gmail.com> wrote:

> In short this would mean (fo me) that if Bob's online computer is compromised,
> let's say with a trojan specialized in adding such a tag, he can create his message

This is not what Ritter (and Zax and me) mean by tagging attacks which can be
carried out by remailers cheating by altering messages that pass through.

On 22 Sep 2021, Stefan Claas <spam.trap.usenet@gmail.com> posted
some

> On Saturday, September 18, 2021 at 4:50:18 PM UTC+2, Stefan Claas
>> On Saturday, September 18, 2021 at 2:24:13 PM UTC+2, Anonymous
wrote:
>> >
>> > > On Friday, September 17, 2021 at 11:01:18 PM UTC+2, Yamn
Remailer
>> > >>
>> > >> "Deterministic header padding to protect against tagging
>> > >> attacks"
>> > >>
>> > >> Can someone please define what that means?
>> > >>
>> > >>
>> > >> https://github.com/crooks/yamn
>> > >
>> > > Hi,
>> > >
>> > >
>> > > https://ritter.vg/blog-cryptodotis-
tagging_attack_on_mixmaster.htm
>> > > l
>> > >
>> > > Regards
>> > > Stefan
>> > >
>> > Thanks, Stefan. That was very thorough.
>> >
>> > Now if I may, back to the original question as the docs are
scant
>> > on details.
>> >
>> > Yamn tries to mitigate tagging attacks by "Deterministic header
>> > padding to protect against tagging attacks"
>> > In layman's terms what does that mean?
>> In my blooming imagination it would tell me that if Alice or Bob
has
>> a Government trojan on their online device (which can't be
detected
>> by AV software) and an exit node is compromised, that TLAs know
>> that they send the message(s).
>
> I tried out the following with GnuPG. I added one visible (should
be
> hidden, like a space character) byte to the end of the checksum in
> an armored message and GnuPG does not give a checksum error.
>
> So, when Bob encrypts offline and transfers to his compromised
> computer third parties would know that he send the anonymous
> and encrypted message and not Alice.

I alluded to this as it applies to photos in an Apple thread.
I can take a totally innocent photo and make it appear as an
undesired / illegal pic in a few seconds. I am certain that I am
not the only one possessing this knowledge. It was acquired from
a support element of an OEM with a vast population of storage
across the planet. There is no doubt that this knowledge is being
abused.

On Friday, September 24, 2021 at 1:45:41 AM UTC+2, anonymous wrote:
> On 22 Sep 2021, Stefan Claas <spam.tra...@gmail.com> posted

> > I tried out the following with GnuPG. I added one visible (should
> be
> > hidden, like a space character) byte to the end of the checksum in
> > an armored message and GnuPG does not give a checksum error.
> >
> > So, when Bob encrypts offline and transfers to his compromised
> > computer third parties would know that he send the anonymous
> > and encrypted message and not Alice.
> I alluded to this as it applies to photos in an Apple thread.
> I can take a totally innocent photo and make it appear as an
> undesired / illegal pic in a few seconds. I am certain that I am
> not the only one possessing this knowledge. It was acquired from
> a support element of an OEM with a vast population of storage
> across the planet. There is no doubt that this knowledge is being
> abused.

Interesting. I read a while ago that it is possibel with ransomware
to encrypt all images on a Canon DSLR, so that the owner of such
a camera can no longer access his photos.

Regards
Stefan

On 9/23/2021 6:45 PM, anonymous wrote:

> I alluded to this as it applies to photos in an Apple thread.
> I can take a totally innocent photo and make it appear as an
> undesired / illegal pic in a few seconds. I am certain that I am
> not the only one possessing this knowledge. It was acquired from
> a support element of an OEM with a vast population of storage
> across the planet. There is no doubt that this knowledge is being
> abused.
>

These are all great and important issues.

But they have nothing to do with the original issue of tagging
attacks and Mixmaster.

On 9/23/2021 3:49 AM, elvis-85792@notatla.org.uk wrote:

> On 2021-09-22, Stefan Claas <spam.trap.usenet@gmail.com> wrote:
>
>> In short this would mean (fo me) that if Bob's online computer is compromised,
>> let's say with a trojan specialized in adding such a tag, he can create his message
>
> This is not what Ritter (and Zax and me) mean by tagging attacks which can be
> carried out by remailers cheating by altering messages that pass through.
>

Thank you.

As I understand tagging, a malicious operator can inject malformed code
in a Mixmaster message that can potentially enable tracking of that
message? Is that correct?

Or can the actual contents of the message be potentially be altered?

On Friday, September 24, 2021 at 3:53:49 PM UTC+2, Anonymous wrote:
> On 9/23/2021 3:49 AM, elvis...@notatla.org.uk wrote:
> > On 2021-09-22, Stefan Claas <spam.tra...@gmail.com> wrote:
> >
> >> In short this would mean (fo me) that if Bob's online computer is compromised,
> >> let's say with a trojan specialized in adding such a tag, he can create his message
> >
> > This is not what Ritter (and Zax and me) mean by tagging attacks which can be
> > carried out by remailers cheating by altering messages that pass through.

Hi elvis,

I am fully aware of this and only liked to point out what else is possible.

BTW. since you use an X-NA Header I cannot reply directly to you.

Regards
Stefan