Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

A novel use for EXIF (etc).

28 views
Skip to first unread message

geoff

unread,
Jan 16, 2023, 6:35:27 PM1/16/23
to

Alan Browne

unread,
Jan 17, 2023, 10:06:32 AM1/17/23
to
On 2023-01-16 18:35, geoff wrote:
> https://www.bbc.com/news/world-asia-china-64206950

Nothing to do with exif.

Steganography uses the image data itself by modifying image bits a very
slight amount up or down to represent the data of the hidden message.
To the viewer of the image there is no apparent information - it is in
"in the noise" so to speak.

With the right s/w and "password" (or key) decoding the data is trivial.

Detecting images with steganography in them is hard as well as it's
literally in the noise. So statistical techniques might find evidence
of it - but still not be able to decode and prove it absent the key and
particular encode algorithm used.

EXIF could be used for additional data, but would be more easily
detected. It could also indicate secretly to the receiver which
specific algorithm was used and which one-time pad key was used - but
this is pushing it a bit.

--
“Donald Trump and his allies and supporters are a clear and present
danger to American democracy.”
- J Michael Luttig - 2022-06-16
- Former US appellate court judge (R) testifying to the January 6
committee

David Brooks

unread,
Jan 17, 2023, 11:24:55 AM1/17/23
to
On 17/01/2023 15:06, Alan Browne wrote:
> On 2023-01-16 18:35, geoff wrote:
>> https://www.bbc.com/news/world-asia-china-64206950
>
> Nothing to do with exif.
>
> Steganography uses the image data itself by modifying image bits a very
> slight amount up or down to represent the data of the hidden message. To
> the viewer of the image there is no apparent information - it is in "in
> the noise" so to speak.
>
> With the right s/w and "password" (or key) decoding the data is trivial.
>
> Detecting images with steganography in them is hard as well as it's
> literally in the noise.  So statistical techniques might find evidence
> of it - but still not be able to decode and prove it absent the key and
> particular encode algorithm used.
>
> EXIF could be used for additional data, but would be more easily
> detected.  It could also indicate secretly to the receiver which
> specific algorithm was used and which one-time pad key was used - but
> this is pushing it a bit.

All very interesting though! 🙂

Can you help with this question, Alan?

https://discussions.apple.com/thread/254563418

--
Kind regards,
David

nospam

unread,
Jan 17, 2023, 11:45:50 AM1/17/23
to
In article <lZzxL.20724$eRZ7...@fx06.iad>, David Brooks
<Davi...@a.loss.now> wrote:

> > EXIF could be used for additional data, but would be more easily
> > detected.  It could also indicate secretly to the receiver which
> > specific algorithm was used and which one-time pad key was used - but
> > this is pushing it a bit.
>
> All very interesting though! ?
>
> Can you help with this question, Alan?
>
> https://discussions.ap

thread hijacking attempt.

Alan Browne

unread,
Jan 17, 2023, 11:58:05 AM1/17/23
to
Colour me shocked.

David Brooks

unread,
Jan 17, 2023, 2:31:28 PM1/17/23
to
On 17/01/2023 16:57, Alan Browne wrote:
> On 2023-01-17 11:45, nospam wrote:
>> In article <lZzxL.20724$eRZ7...@fx06.iad>, David Brooks
>> <Davi...@a.loss.now> wrote:
>>
>>>> EXIF could be used for additional data, but would be more easily
>>>> detected.  It could also indicate secretly to the receiver which
>>>> specific algorithm was used and which one-time pad key was used - but
>>>> this is pushing it a bit.
>>>
>>> All very interesting though! ?
>>>
>>> Can you help with this question, Alan?
>>>
>>> https://discussions.ap
>>
>> thread hijacking attempt.
>
> Colour me shocked.

Try this one then:-

https://discussions.apple.com/thread/252848419

nospam

unread,
Jan 17, 2023, 2:36:33 PM1/17/23
to
In article <tq6t2a$3c4r5$1...@hunterbd.eternal-september.org>, David
Brooks <Dav...@E-S.invalid> wrote:

> >>> All very interesting though! ?
> >>>
> >>> Can you help with this question, Alan?
> >>>
> >>> https://discussions.ap
> >>
> >> thread hijacking attempt.
> >
> > Colour me shocked.
>
> Try this one then:-
>
> https://discussions.ap

try foad.

Carlos E.R.

unread,
Jan 17, 2023, 2:42:15 PM1/17/23
to
Another thread hijacking attempt.

--
Cheers, Carlos.

Carlos E.R.

unread,
Jan 17, 2023, 2:42:15 PM1/17/23
to
On 2023-01-17 00:35, geoff wrote:
> https://www.bbc.com/news/world-asia-china-64206950
>
> geoff

There is no mention of "EXIF" in the article.

--
Cheers, Carlos.

geoff

unread,
Jan 17, 2023, 3:48:13 PM1/17/23
to
On 18/01/2023 8:39 am, Carlos E.R. wrote:
> On 2023-01-17 00:35, geoff wrote:
>> https://www.bbc.com/news/world-asia-china-64206950
>>
>> geoff
>
> There is no mention of "EXIF" in the article.
>

Not specifically.

¿ Y que ?

geoff

Carlos E.R.

unread,
Jan 17, 2023, 4:06:25 PM1/17/23
to
Well, I want to know what is that novel use for EXIF claimed in the
subject line by you.

--
Cheers, Carlos.

Alan Browne

unread,
Jan 17, 2023, 4:53:50 PM1/17/23
to
Because EXIF is not at all the method used. At all.

See my other post about steganography ... which /is/ mentioned in the
article. Because that is a pretty nice way to do this sort of thing in
the open with a low probability of detection.

David Brooks

unread,
Jan 17, 2023, 7:16:12 PM1/17/23
to
On 17/01/2023 21:53, Alan Browne wrote:
> On 2023-01-17 15:48, geoff wrote:
>> On 18/01/2023 8:39 am, Carlos E.R. wrote:
>>> On 2023-01-17 00:35, geoff wrote:
>>>> https://www.bbc.com/news/world-asia-china-64206950
>>>>
>>>> geoff
>>>
>>> There is no mention of "EXIF" in the article.
>>>
>>
>> Not specifically.
>
> Because EXIF is not at all the method used.  At all.
>
> See my other post about steganography ... which /is/ mentioned in the
> article.  Because that is a pretty nice way to do this sort of thing in
> the open with a low probability of detection.

Will running 'First Aid' in Disk Utility identify/erradicate
malware hiding in a Mac operating System?

--
Kind regards,
David

Carlos E.R.

unread,
Jan 17, 2023, 7:57:25 PM1/17/23
to
What the eff does that have to do with what we are talking about?

--
Cheers, Carlos.

nospam

unread,
Jan 17, 2023, 8:09:24 PM1/17/23
to
In article <3hsj9jx...@Telcontar.valinor>, Carlos E.R.
absolutely nothing.

geoff

unread,
Jan 17, 2023, 10:13:20 PM1/17/23
to
No reason why info related to the picture could not be easily added to
EXIF metadata. Yes, smaller capacity and not as sophisticated as the
case in point.

But as I said, a novel use for EXIF data.

geoff

geoff

unread,
Jan 17, 2023, 10:15:19 PM1/17/23
to
On 18/01/2023 10:53 am, Alan Browne wrote:
> On 2023-01-17 15:48, geoff wrote:
>> On 18/01/2023 8:39 am, Carlos E.R. wrote:
>>> On 2023-01-17 00:35, geoff wrote:
>>>> https://www.bbc.com/news/world-asia-china-64206950
>>>>
>>>> geoff
>>>
>>> There is no mention of "EXIF" in the article.
>>>
>>
>> Not specifically.
>
> Because EXIF is not at all the method used.  At all.
>
> See my other post about steganography ... which /is/ mentioned in the
> article.  Because that is a pretty nice way to do this sort of thing in
> the open with a low probability of detection.

But there is no reason that I could not include "NNNN is a dick" In
EXIF. Which is the point I was making, and not the point the article was
making.

geoff

David Brooks

unread,
Jan 18, 2023, 2:32:30 AM1/18/23
to
A bright fellow like you, Carlos, could have Googled that sentence and
ended up here:-

https://discussions.apple.com/thread/254546074

Do you agree with the responses given?

TIA

David

Carlos E.R.

unread,
Jan 18, 2023, 4:53:11 AM1/18/23
to
Brrrr.... :-(


--
Cheers, Carlos.

Carlos E.R.

unread,
Jan 18, 2023, 4:53:11 AM1/18/23
to
That is something that is not mentioned at all in the article, probably
because it is not true.

--
Cheers, Carlos.

Carlos E.R.

unread,
Jan 18, 2023, 4:53:12 AM1/18/23
to
I am not even going to read that link. You are constantly trying to get
people to read that.

Go an start a thread of your own in the appropriate Usenet group for
that pet subject of yours.


--
Cheers, Carlos.

David Brooks

unread,
Jan 18, 2023, 7:30:05 AM1/18/23
to
Why are you so grouchy? Have you got Spanish Flu?

> Go an start a thread of your own in the appropriate Usenet group for
> that pet subject of yours.

I /have/ mentiond it on uk.comp.sys.mac. Perhaps you will suggest others?

What do you consider my "pet subject" to actually be, Carlos?

--
Kind regards,
David

Alan Browne

unread,
Jan 21, 2023, 6:00:28 PM1/21/23
to
I suggest you sharpen your points before making them in the future -
esp. when linking an article in the post that describes something else.

geoff

unread,
Jan 22, 2023, 5:57:07 AM1/22/23
to
On 22/01/2023 12:00 pm, Alan Browne wrote:
> On 2023-01-17 22:15, geoff wrote:
>> On 18/01/2023 10:53 am, Alan Browne wrote:
>>> On 2023-01-17 15:48, geoff wrote:
>>>> On 18/01/2023 8:39 am, Carlos E.R. wrote:
>>>>> On 2023-01-17 00:35, geoff wrote:
>>>>>> https://www.bbc.com/news/world-asia-china-64206950
>>>>>>
>>>>>> geoff
>>>>>
>>>>> There is no mention of "EXIF" in the article.
>>>>>
>>>>
>>>> Not specifically.
>>>
>>> Because EXIF is not at all the method used.  At all.
>>>
>>> See my other post about steganography ... which /is/ mentioned in the
>>> article.  Because that is a pretty nice way to do this sort of thing
>>> in the open with a low probability of detection.
>>
>> But there is no reason that I could not include "NNNN is a dick" In
>> EXIF. Which is the point I was making, and not the point the article
>> was making.
>
> I suggest you sharpen your points before making them in the future -
> esp. when linking an article in the post that describes something else.
>

Jeeze there are some pedantic fucks around here.

geoff

Carlos E.R.

unread,
Jan 22, 2023, 7:03:40 AM1/22/23
to
If we were pedants, we'd tell you that pedantic is not the correct word.

--
Cheers, Carlos.

Alan Browne

unread,
Jan 22, 2023, 10:06:19 AM1/22/23
to
A lot of lazy writers too....

Alan Browne

unread,
Jan 22, 2023, 10:06:51 AM1/22/23
to
hee-hee.
0 new messages