Peter Bourgon

> There lies the risk. By default, most strings in most languages have comparison operators that are case sensitive. This means that these 3 UUIDs, despite being identical in their...

Wow! You learn something new, and frequently terrible, every day.

Yeah, there's a lot to consider. I'm definitely against an interface with more than 1 method, see [this presentation from a Go kit collaborator](https://go-talks.appspot.com/github.com/ChrisHines/talks/structured-logging/structured-logging.slide#1) for a good recap of my...

You're still on major version 0, it shouldn't be a problem.

"Data race" is unfortunately not a well defined term. Broadly, there are two classes of data races. One class is totally safe, but somewhat reliably produces nondeterministic behavior. Another class...

Your distinction is noted and understood, but it not well-understood by the industry. > If the "data race/race condition" terminology is not clear to the majority of people (which is...

In fact, dropping the phrase "race condition" from the document entirely would be ideal.

Also, I apologize for the way the comment in the wiki page was phrased. It was imprecise and, as you suggest, misleading in its own way. I've corrected it, even...

> The metrics I would like to send to somewhere in JSON is defined as [Ethereum consensus client metrics](https://github.com/gobitfly/eth2-client-metrics) and [Beacon chain metrics](https://github.com/ethereum/beacon-metrics/blob/master/metrics.md). This is out of scope for this...

ULIDs and UUIDs are both 128 bits worth of data but they are not compatible with each other. ~More specifically: all ULIDs are UUIDs, but not all UUIDs are ULIDs....