update package.json

[daviddias]

Created this PR to make travis run again on clean master to check the errors that @nginnever mentioned

[daviddias]

yeah, this is weird: https://travis-ci.org/ipfs/js-ipfs/jobs/128359184#L2133-L2136

[daviddias]

@dignifiedquire you might want to take a look. This was introduced with the update on libp2p-swarm: libp2p/js-libp2p-switch@85a0647

Probably the problem was always there, but now it is really easy to see

[daviddias]

18:57 daviddias: everything before this commit is passing for me
18:57 1251dea

[nginnever]

18:57 daviddias: everything before this commit is passing for me
18:57 1251dea

I take that back, I started hard reseting commits again one at a time and the commit before this one threw the spdy error as well.

[daviddias]

@nginnever wanna check with libp2p-swarm 0.12.5 ?

[nginnever]

Downloaded the archived libp2p-swarm v0.12.5 and npm linked it to the js-ipfs repos, seems to fix the issue.

[nginnever]

Changed libp2p-swarm to throw transport close errors

async.each(
  Object.keys(this.transports),
  (key, cb) => this.transports[key].close(cb),
  callback
)

and received

Error: there are no listeners

In js-ipfs libp2p core swarm.close

[daviddias]

I've noticed that on Travis, the "isDaemonOn" always yielded true, making the 'cli offline' tests work online. Checking if this was the issue.

Good best practice, never do instanceof for 'types' outside the scope of the module, cause different versions will have different signatures and you never know which version will be thrown at the func.
@nginnever @dignifiedquire

[daviddias]

Yep, it fixed the 'weird fails' on the block tests : )

[nginnever]

I put these changes into the files PR.

[daviddias]

Hopefully, this will give us a better answer spdy-http2/spdy-transport#26 (probably the right answer is the one that is most simply, we need to check for .on('error') in every muxer we create

[daviddias]

GREEN :D