-
Notifications
You must be signed in to change notification settings - Fork 30
Make it official that mplex is actually a thing #68
Comments
Nothing against? Otherwise, I'll just go ahead :) |
At this point, I'm pretty sure it has converged back. Is this not the case? |
What do you mean by that? |
As in, don't both implementations now follow the "multiplex" spec? Wait, which came first, the javascript version or the go version? |
First -> http://npmjs.org/multiplex, Then -> go-multiplex (impl without spec, Then new js-multiplex (built by us, still no spec), Then -> interop achieved. |
So, looking at that code, I don't see much divergence. |
@Stebalien are you sure you are comparing the right codebases? https://github.com/dignifiedquire/multiplex vs https://github.com/maxogden/multiplex ? Also, the point I'm making here is that today we today use a muxer which is a fork of a fork of a muxer and none has a spec, which makes it really hard to track changes and to have a conversation about how it should work. Note, the multicodec is already |
There's divergence in terms of code but the I don't see any obvious protocol-breaking changes. However, without an original spec, it's hard to tell. Also, I've now put together an up-to-date reverse spec here: https://github.com/whyrusleeping/go-multiplex/blob/master/spec.md Yes, I agree that calling it |
Done with #72 |
The multicodec is
mplex
, the implementation has diverged greatly frommultiplex
and currently there is no spec anyway.I propose:
js-libp2p-mplex
Me and @Stebalien have been chatting about nuances of multiplex and its feature set and there are a lot of
?
,ugh
andmeh
. Another great example is the CR from @whyrusleeping on this PR libp2p/go-mplex#12 (comment) :)The text was updated successfully, but these errors were encountered: