I've been very happy with Terraform. I also have a huge investment in Cloudformation and it doesn't take long to discover that Cloudformation isn't implemented consistently across services, it's full of bugs, and unless your bug happens to impact Amazon internally you can wait years for a fix. New functionality for existing resources can also take years to materialize. Back when I had an eight figure AWS budget they would offer to let me fund a Cloudformation fix (seriously! They wanted me to pay them to fix their errors) but since moving to a smaller shop they don't even offer that.

With Terraform (and its AWS provider) I've found things are generally implemented consistently, I've not encountered any crashing bugs yet, and all of the manual steps I've had to do with the AWS CLI have Terraform equivalents.

Looking at what CloudFormation can and can't do it seems like completely different teams work on a service and its CloudFormation support. I guess Amazon use the AWS API internally?

They've been publishing a lot of the alleged source code to their CFN providers of late, so one can see how the sausage is made: https://github.com/aws-cloudformation?q=aws-cloudformation-r...

I say alleged because who knows if that's the code that's really running. And related to this current thread even if it is the code, with this being Amazon who knows if I ever found a bug and PR-ed it that it wouldn't /dev/null for 5 years before being closed by some stale-issue-bot or something

I work at AWS. The service teams do own and write their CloudFormation providers, although if you've written a custom provider you'd see it is somewhat clunky, so it's sometimes considered more of an operational burden, and you can tell.

We dogfood both the SDK and CloudFormation internally, we just deal with its numerous gripes much the same way you would externally (although we can also contact service teams directly if needed).

Million dollar question would be if all the “feature requests” that AWS support opens when they don’t know what else to do actually go someplace or if they just get swept aside every few days. I’ve always been under the impression they receive very little or no attention. Even bugs where I have included a reproducible test case that results in an internal error, they just keep trying to close the ticket every few days until you go on vacation or e-mail burps and you miss the notice.

I work for AWS. We have an internal tracking system containing customer feature requests, and our service teams review them biweekly and they get prioritized by product management. “Working backwards from the customer” is gospel here.

That is cool to know! For several years I have been tell the support people not to bother with feature requests because I didn’t think they went anyplace, maybe I’ll let them open them going forward.

Really hoping they go public. They are one of two tech companies I would invest in.

I'm assuming that there's a communications lockdown and freeze ahead of an IPO, and that extends even to reviewing pull requests. But I know jack and perhaps I'm just seeing ghosts.

What's the second?

Datadog. I've used their products for years and I believe in their team. Picked most of what I have up after their ipo. Hashicorp is similar (but I haven't been a paying customer of theirs).

I'm a timid investor and am pretty nihilistic about tech in general. I forget the exact Charlie munger quote, but it was something about staying in your circle of competence.