Update process to ~0.11.0

[zertosh]

Fixes #1179. This also solves my major-version-bump dilemma from #1230 because this merits it for sure.

The reason for having to pass in clearTimeout into the context in tests is because there is now a queue in process.nextTick that depends on it (see defunctzombie/node-process#38 and defunctzombie/node-process@v0.10.1...v0.11.0.).

cc: @jmm @terinjokes

[zertosh]

cc: @slorber

[terinjokes]

Glad to get the Q issue resolved. I'm happy to merge both these in and do a major release. @jmm?

[zertosh]

@terinjokes Cool! I merged browserify/insert-module-globals#43 and bumped it here in 064f5a1 - just for consistency, doesn't really affect anything in browserify.

[slorber]

nice is there a browserify release yet?

[terinjokes]

@slorber not yet ✨

[jmm]

@zertosh @terinjokes You guys probably understand this issue better than I do, but I looked through a bunch of the related issues / PRs and it LGTM.

This also solves my major-version-bump dilemma from #1230 because this merits it for sure.

Does this break BC somehow? I got the impression that stuff that's already working would continue working with this change. Or would the changes in defunctzombie/node-process/pull/38 actually break something people were relying on before? The other stuff I saw in recent commits there didn't seem like it would be breaking. Sorry if I'm being dense here.

Aside: it sounds like maybe the issue with Q in particular has been addressed independently there -- not that it has any bearing on this PR.

[zertosh]

@jmm I would think this is a BC because it fundamentally changes how process.nextTick works by enforcing a queue that doesn't die on errors (plus, now it depends on clearTimeout, which broke our tests). Granted, it is fixing a bug, but a bug that might've been relied on. Since this is such an obscure feature - does that kinda fall into the semver gray area? I don't know :( I could really use some guidance right now.

[jmm]

@zertosh I'm not an expert on semver, so it does seem like a gray area to me. If you think people may have relied on the previous behavior then I think it would make sense to bump the major version. I wonder what node-process would've done with the version if it had already been >= 1.

[terinjokes]

Reading more into the issue for the changelog, I'll find it really strange if anyone did rely on the behavior, since it prevented any future process.nextTick tasks from running.

[zertosh]

@terinjokes Maybe? I'm by no means a semver expert 😛