amaurymartiny
Joined 9 years ago
Joined 9 years ago
I'm using a similar technique to check deliverability, i.e. parsing SMTP responses.
I released the open-source lib some years ago, and built a Saas only this year in April. It started generating revenue (~$100MRR) since September, and I'm not really touching it anymore.
Using Tor, I can verify around ~60% of emails. Notably, the exit nodes can connect to Gmail/GSuite servers. So it's not that bad.
If Tor's exit nodes are blocked by a server, I fallback to Heroku, where I have right now 3 fixed-IP instances. And if I see that one of these also gets blacklisted, I would have it self-destruct and relaunch, Heroku assigns a new IP on each new instance.
The tool connects to the target email's SMTP server, and parses the response of the "RCPT TO" command. To avoid IP blacklisting, I use Tor.
The code is in Rust, 100% open-source. The core library is here: [link redacted]
I believe cognitive cost would also be a good denominator. Something like "we find that an increase of 10 µg/m³ raises the probability of a chess player making an error by 1.5 percentage points. The same 1.5 percentage points would be caused by smoking 2 cigarettes per day."
Plug: I'm the author of an open-source mobile app[2] which does this pollution-cigarettes conversion. I think cigarettes gives a strong image of air quality for those who are unfamiliar with AQI.
[1]: http://berkeleyearth.org/air-pollution-and-cigarette-equival... [2]: https://shootismoke.github.io/