Alex Dubov

I'd say, this stuff should be added to the docs near the beginning (in some sort of best practices section). After all these years I suddenly learn that we don't...

We have further investigated the Aurora issue and apparently, this is not a bug, but a side effect of a feature called "zero downtime patching" used by AWS to apply...

Happens with AWS Aurora as well. With Aurora MySQL too, even though it's a different driver. Kind of makes one think it's Go TLS issue.

Of course, I don't want this to be about thread local, it's a long and difficult topic (just mentioned it for the context). All I personally need is exported and...

I know about the `go:linkname` hack. Everybody who's doing the sort of thing I enumerated in my OP does. :-) My point here that time has come to make it...

We can either make sharded values with fast random sources or we can make fast random sources with sharded values. Sort of like chickens and eggs. :-)

What's wrong with defining user visible `rand.Uint32()` in terms of internal `runtime.fastrand()`? No fancy stuff, just that and call it a day. This is the de-facto situation today with the...

(Original comment replaced with this. Sorry.) I have went on to see what people do on Github in general. Right now, there are several common techniques to obtain fast randoms:...

Somehow it feels that newly added `maphash.MakeSeed` could solve the present problem. The users are allowed to call it at will affecting the generator state, but god forbid mere mortals...

"it should be possible to make the global source lock-free" How creating another level of PRNG indirection in some unspecified future is any better than simply exposing `fastrand`?