Matt Way

I'd really like to stop introducing pointer values in config, especially for scalar types. Maybe we could start using a consistent convention of "0 for disable, -1 for default", or...

> @robskillington & @mway, will you please review or tag the persons who can review? > My apologies for bothering you. No problem, thank you for the PR! At a...

> @mway thoughts on this? Expanding the API scope is probably not ideal, wonder if there's a similar way to support this without creating another interface. agreed - we want...

We haven't evaluated revive internally yet (to my knowledge, anyway), though we're aware of it and on paper it does look promising. We don't recommend things that we don't do/use...

It's true that there's not really a "one-size-fits-all" approach, but I don't think the issue is whether the send is sync/async, or whether channel usage is error-prone: the issue that...

> The key points here being the first two They're good points overall. Like anything, there are exceptions, but in common cases these hold true. This is effectively the spirit...

> Sometimes goroutines can block indefinitely. And it's not a rare situation: the Go runtime is optimized for exactly this situation. So it's not necessarily a problem. Right? This is...

On second thought, maybe this is either the same as or a subset of #67. Could be worth combining them.

Thanks Prashant! The point is valid, though I think this is probably a bit niche compared to how (more or less) broadly applicable the other guidance is. Based on how...

Similar comment to #124: We haven't evaluated revive internally yet (to my knowledge, anyway), though we're aware of it and on paper it does look promising. We don't recommend things...