[HN Gopher] Debugging an empty spam email (2016)
___________________________________________________________________
Debugging an empty spam email (2016)
Author : psobot
Score : 25 points
Date : 2022-06-15 00:17 UTC (22 hours ago)
(HTM) web link (blog.petersobot.com)
(TXT) w3m dump (blog.petersobot.com)
| mike-cardwell wrote:
| Shoulda used https://www.parsemail.org
| GrumpyNl wrote:
| So it turns out that the spammer doesnt know what he is doing by
| spamming unreadable mails.
| trinsic2 wrote:
| I'm not sure if you read the article, but my take on it was
| that the spammer hid a scam message in a embedded PNG file that
| may get decoded on older style email clients, but gmail did not
| render the data-uri for some reason. It might have been part of
| gmails spam recognition to not render the image in this
| instance. The message itself got through the spam filters, but
| it rendered empty because the embedded image did not get
| displayed.
| kingcharles wrote:
| I'm unsure why Gmail didn't render the image. It sure as hell
| renders all of the ones I get if the email itself is not
| flagged as spam (as in this case). The image is embedded into
| the email, so there is no privacy or tracking issue from
| rendering it.
| ljp_206 wrote:
| As soon as you suggested that such an embedded PNG may only
| work on an older client, a lightbulb went off for me.
|
| Most spam is predicated on attacking those too
| technologically literate to vet the attack. A smart scammer
| could create an email that would go under the radar of more
| modern client users, who are likely to report the message as
| spam and reduce the scammer's reach. If the message only
| works in older clients used by softer targets, then their
| chances of success are increased.
|
| If this is the intended method of the attack, it's quite
| clever. Imagine if you could still blanket spam every email
| you come across, but only target users with old, outdated
| clients, who are likely older, less technologically savvy,
| etc... It would be well worth the R&D time.
|
| Such reasoning follows the theory that spam messages include
| many typos in order to weed out 'smart' users, who are not
| desirable targets anyway.
| logifail wrote:
| > likely to report the message as spam and reduce the
| scammer's reach
|
| Q: Is reporting spam (and if so, to whom?) actually of
| benefit? If so, how?
|
| If you "report" spam at Gmail or in Office 365, _what
| actually happens_?
|
| Does some magic kick in and move - in real time - all
| identical (similar?) messages out of Inboxes and into Spam
| folders across the globe? Or is it more about training
| filters for the next wave?
| 35mm wrote:
| > Does some magic kick in and move - in real time - all
| identical (similar?) messages out of Inboxes and into
| Spam folders across the globe?
|
| Not exactly. But the domain they are sending from will
| get added to shared blacklists.
|
| Spammers don't send all in one go, so that same message
| going out to the next batch of people could be stopped.
| logifail wrote:
| > Not exactly. But the domain they are sending from will
| get added to shared blacklists.
|
| In real time? After one report or are multiple reports of
| a spammy domain required?
|
| I use blacklists (and indeed greylisting) on my personal
| email servers but I'm curious how much we know about how
| MSFT/O365 and Gmail handle this stuff.
___________________________________________________________________
(page generated 2022-06-15 23:02 UTC)