Miguel Ángel Ortuño

> Would be awesome to have a way to reuse them when the item is not needed anymore. Definitely agree. I think this is a point of improvement that could...

When you get a chance, take a look at these changes: https://github.com/ortuman/gomemcache/commit/8bb94a59999fc52836308966af8ec058d3d706f0 Also I've done some benchmarking and seen a slightly better result when reusing buffers. I guess the true...

> The benchmark is wrong they both make the slice. The first approach does not implicitly invoke `Close` method, whereas the second one does. This causes no buffer object to...

Oh! You're 100% right. Just changed it @ https://github.com/ortuman/gomemcache/commit/ab4be5bd29b01f1f7687db3b4333890a514ac6a2

Sure thing! As you mentioned I just wanted to experiment on my own repo before doing a formal fork. If this goes on I'll follow your suggestion. Thanks!

👋 > I'm unsure if this is a bug in jackalctl allowing invalid user input or Jackal using the incorrect username. That's the key. The `jackalctl` command only expects username...

Sweet @SamWhited! 👏 👏 👏 > I can't really test this locally (as the script also appears to operate in GOPATH mode which hasn't been used for a while). The...

I see. I forgot I had `GOPATH` variable set (a legacy from the old days). 🤦‍♂️ I wonder if we should follow a similar approach as the `go install` command....

> Do you have a reference for what makes this safe? E.g. is it stated in the gRPC documentation? There's no explicit statement about it (haven't found it at least)....

> I think that millisecond scale is not as neglible as it can be micro/nanosecond, can we try understand where that increase comes from? One of the things I'd blame...