Matthew Riley

Agree that #1754 looks like a good path forward for resolving this issue. The "Shell" in `ShellExecute` is the [Windows shell](https://en.wikipedia.org/wiki/Windows_shell) (GUI), i.e. Explorer. Not `command` or `cmd`. That said,...

@ewindisch Do you folks have documentation or guidance for what constitutes an embargo-class vulnerability? This isn't remote, unauthenticated RCE. The danger here doesn't seem nearly acute enough to justify a...

Yep, I was wrong to say this wasn't RCE. It's a bad bug, and a great testament to why robust image signing is a great idea. However, exploitation isn't trivial:...

Perhaps it's more accurate to say: these tests have "golden values" that rely on the X25519 implementation in this codebase, which doesn't appropriately decode keys before using them. Other X25519...

That's probably a bit extreme. This codebase (and others by WhisperSystems) normalize keys on _creation_ rather than on _use_. In practice, this only causes difficulty when importing keys created by...

I wondered if this might relate to health checks and healthy intervals. Here's update events for a `cloud.Service`-backed ECS service:  here's the accompanying graph of "Healthy Hosts" from the...

The two new tasks are created at 19:18:29. By 19:18:36, both tasks have logged `Serving PPC API on [::]:8080`.

I happened to catch a service deployment while investigating this issue. Here are the deployment events:  and here's the load balancer target listing as of 13:03:28 PDT:  Meanwhile...

Unanswered question where someone else sees similar behavior: https://stackoverflow.com/questions/47256085/aws-network-elb-take-4-minutes-to-recognise-target-as-healthy

Filed a support request to ask AWS what causes a target to move from `initial` to `healthy`.