| FazBrowse GitHub Viewer | Trending | | Home |
| Tools: [Download Repo ZIP] [Original HTTPS Page] |
child.send() returns undefined, contrary to the docs. Update docs and remove related dead code.
Sorry, something went wrong.
There was a problem hiding this comment.
This is not a good change. I remember this was introduced deliberately because unbound queue growth was a real problem with some applications.
Sorry, something went wrong.
There was a problem hiding this comment.
It's currently an unused value returned from _send() and swallowed by send() and has been since 0.12.0. The last time end users were able to get a boolean returned from send() was 0.10.40. I'm not arguing that you're wrong. I'm just providing additional info/context in case that changes your estimation of the code's importance. (I'm guessing it does not and I certainly value your judgment on it more than my own. You have internalized a ton of historical context I am oblivious to.)
Sorry, something went wrong.
|
This would be a semver-minor at least if this lands which takes it out of the v4.x queue... |
Sorry, something went wrong.
|
@jasnell This does not change the behavior of Node. That return statement that is removed is basically dead code. The return value (from _send()) is unused and discarded by send() and the end user never sees it. Node/iojs have been returning undefined from send() in all cases since at least iojs 1.0. EDIT: That said, I'm fine with it not landing in LTS. And it looks like this isn't going to land anyway and that @bnoordhuis (and probably others?) favor #3516 (which changes the code to use that value) over it anyway. All of which is totally A-OK by me! |
Sorry, something went wrong.
| Back | FazBrowse Home | New Git URL |
child.send() returns undefined, contrary to the docs. Update docs and
remove related dead code.
This is an alternative to #3516. That PR updates the behavior of the code to reflect the docs. This updates the docs to reflect the behavior of the code.