Email tracking pixel support | Hacker News Reader

Email tracking pixel support

rant.gulbrandsen.priv.no

23 pointsakakievich14 comments

Why are tracking pixels still even a thing. Wouldn't this all be solved is the top 3 mail providers would just decide to hit all embedded resources at the moment a message is received and proxy or even cache it after? This would render this method ineffective for the vast majority of receivers

In order to proxy/cache it they would have to rewrite https resources in user mail to point at their servers. Technically possible but a PR nightmare for little gain to them.

They do this already. Similar to Github’s Cameo service for proxying readme images.

Wow, I just checked and yep there it is - google URL! I wonder if it caches automatically or only when read. If the latter I bet you could game the system by sending a tracking pixel with a unique source URL. Will have to stand up a test server and see what resource gets requested when.

It varies by the reputation of the domain. Most images get hit once when the email is initially opened, and cached basically forever. If you send tons of tracking pixels to multiple users you'll eventually see the images being loaded and cached as soon as you deliver the message.

I thought google started doing this years ago https://gmail.googleblog.com/2013/12/images-now-showing.html

They might be proxying it, but it still only gets hit once you open the mail as I can remember after sending myself a testmail a while back, embedding an image and watching the webserver logs.

Only if you enable images for the email.

If you don't enable images, the tracking pixel is never retrieved.

Capital One uses email tracking to 'verify' your email address. I get a yearly email from them saying I've not read any of their email but I disable downloading images. These technical hacks for business purposes are lousy.

This is funny and wonderful.

However, if I can take the joke way too seriously for just a second:

Doing this with emails will just cause your target to think something is messed up on your end. I'm not sure it would have the desired effect unless a lot of people were doing it.

In theory, a better strategy would be to inject the font into an environment where users are primed to assume that if something goes wrong it's the environment's fault, not the publisher. For example, a website.

I wonder how hard it would be to get an Open recreation of Smelvetica onto some kind of Smelvetica-specific CDN that everyone could share, and then to get that CDN url added to EasyList?[0]

Would anyone be likely to install an adblocker over that, or is it more likely that they would just blame the website?

[0]: https://easylist.to/pages/development.html

Mirror because it's obvious parody, which is protected, also because anarchy: http://archive.is/zeGxt

You can also just clone the repo and checkout something from the history.

That was the most amusing part of the article. They sent Smelvetica a takedown notice and were satisfied by a quick "git rm <files>". You can still get at the target of the takedown notice through the web interface; https://github.com/tholman/smelvetica/blob/7186794b56c365e6f...

ironic because archive.is has insidious tracking pixels.

*.pixel.archive.is

If you want to Smelvetisize other fonts, I wrote up a quick script and how-to do it with python. https://readevalprint.com/Schmelvetica.html

Reply on news.ycombinator.com