Post #1737989
2026-04-13 18:16 UTC
Setting up a brand new AWS account, when all you need for email is an instance running postfix: "Hi AWS support, I need port 25 unblocked, yes I know email good practices and can spell SPF, DKIM, and DMARC" -> \*approved\*
When you plan to need SES: "Hi AWS support, I need access to SES, here's a detailed description of my project, here's the types of email I plan to send and expected volume, here are some example emails (insert "confirm your email address" etc type notification text), here's our policy about handling unsubscribes and complaints, here's an outline of what we plan to have in our privacy policy, here is a blood and urine sample, I hereby swear allegiance to Jeff Bezos as the supreme overlord of cloud computing, etc etc etc" -> \*denied\* "try giving us more money for a few months then try again" -> (internally) "fuck it, sendgrid it is, you just lost a customer"
Because all projects are able to scale their infrastructure spend over several months before they ever need to send a fucking email, right, apparently? and nobody ever starts on long-leadtime things that need support requests first?
Replies (1)
-
@SiteRelEnby@transfem.social 2026-04-13 18:58
The stupidest thing is AWS has MORE visibility into outbound email with SES than they fucking do with just spinning up postfix on EC2 where the TLS session is betwen the fucking instance and the remote server, as opposed to SES being in the middle and able to see the fucking plaintext.