That has to be the understatement of the century.

Or if Excel just said, Sorry, you can't use that formula with this formula? Or with these types of numbers, or this shape of data, etc?

I guess I was meaning less the user-facing stuff, and more the underlying architecture, the uptime, the general availability. Does the thing do what you expect it to do? Does it work? Is it on? Yes, for the most part it does what it says on the lid very well.

Multiple account blocks with no warning, no questions, over weekends when no one sees it. False positives triggering automated actions from their loosey string-matching for at-best guesswork bad URLs. No recourse. No communication. Dark(est) patterns to try and contact support. Or just no support. No response. Repeated global blocks for the same exact reason that was already solved.

Here's my favorite example: There is a massive service called Formaloo, which is exactly what it sounds like: a copy-cat of Google Forms, Microsoft Forms, every other generic form-builder service. One of our customers included their Formaloo link, which opens a quick "exit interview questions for graduating students from a private middle school". We send 60,000 emails per day and have a 99% reputation IN Sendgrid for 10+ years. With no warning, Sendgrid blocks our entire account because the word "formaloo" can be used for spam. So for 4 days with absolutely no communication, no response, no indication which specific email was at fault. We called, we emailed, we tried chat, we emailed more, we responded to tickets, with no response.

Guess what finally unblocked our account after 4 days of silence? We called their enterprise sales number, and it was fixed within minutes.

This is the third time we've been held hostage for multiple days with no recourse, no review, no "hey, we noticed this link, can you check on this before we block the account", or "hey, we quarantined this one email for your review while we let the other 60,000 go through, is this legit?" It goes from "everything's fine" to "we're screwed" in a heartbeat. And then we have to pay them $1000/month still! No way.

Needless to say, we are dropping Sendgrid.

But we have no issues whatsoever with Azure Devops....ever. It's excellent. Seriously.

Does anyone know why the experience between Github and DevOps is so different, if they're supposed merging the two? Or at least seemingly related? Or are they not at all?

Or is it simply because Azure is "enterprise" and Microsoft cares about that more?

Sendgrid, on the other hand, I would avoid like the plague.

And it has turned into a decent chunk of business over the long run.

And I can't think of anything to add either.