bigs

thanks y'all! edit: and i'll clarify, i've definitely got some experience with it, but not much more than i got out of undergrad. excited to dive deeper.

I'm experiencing this issue as well. We're using reCAPTCHA v2 in "invisible" mode. The error does not occur on the latest Chrome (on Ubuntu 18.0x), but is a show-stopper in...

@yusefnapora the slice definitely plays nicely with our [metrics](https://github.com/libp2p/go-libp2p-core/blob/master/metrics/register.go) package. you can always iterate the values and build the slice yourself. it's equivalent to a `Values` method anyway.

hmm can you give me steps to reproduce this or propose a patch? definitely down to fix this, just not fully sure i understand

Going to check this out first thing tomorrow. Thanks and sorry for the delay, lost this one in the other PRs.

@raulk @Stebalien just rebased this and updated the branch to reflect the core refactor, amongst other things. tests passing, this should be ready for a review.

per @diasdavid's suggestion should we close this temporarily and move to an RFC? edit: i actually mean move it to a discussion and specification session in specs?

Hopping back on this thread... I just did some work with this library for a little demo project of `libp2p` and found I had to do quite a bit of...

@cretz That looks lovely! Would you have any interest in implementing a [libp2p `Transport`](https://godoc.org/github.com/libp2p/go-libp2p-transport#Transport)? Edit: Keeping in mind that this interface will shift a bit soon!

Definitely a fan of the per-protocol limiting. Perhaps this could be handled with a weighting system? Assign weights to protocols and then set global settings (i.e. throttle after this amt...