feat: automatically enable allowHTTP1 by default with http2 & hmr#186
feat: automatically enable allowHTTP1 by default with http2 & hmr#186illusionalsagacity wants to merge 6 commits intoshellscape:masterfrom
Conversation
shellscape
left a comment
There was a problem hiding this comment.
I see the new test fixtures added, but not any tests. Is that pending or uncommitted?
README.md
Outdated
| _Note: If the build process generates errors, the client (browser) will not be notified of new changes and no HMR will be performed. Errors must be resolved before HMR can proceed._ | ||
|
|
||
| _Note: If using in combination with `http2`, the `http2` option `allowHTTP1` must be enabled for the HMR WS connection to work._ | ||
| _Note: If using in combination with `http2`, the `http2` option `allowHTTP1` will enabled by default for the HMR WS connection to work._ |
There was a problem hiding this comment.
| _Note: If using in combination with `http2`, the `http2` option `allowHTTP1` will enabled by default for the HMR WS connection to work._ | |
| _Note: If using in combination with `http2`, the `http2` option `allowHTTP1` will be enabled by default for the HMR WS connection to work._ |
There was a problem hiding this comment.
Yeah, pending I will add another case for it.
There was a problem hiding this comment.
So I can't get the integration tests to run to completion locally, particularly the multi-hmr one seems to be troublesome. It does happen as well on the master branch, so I'm not quite sure what is going on there.
There was a problem hiding this comment.
Can you tell more about the test issue? How does it fail exactly?
|
@illusionalsagacity I'll try and step into your branch to get the tests work. probably later this week. |
Alright, thanks. To answer @bebraw the issue I'm seeing is that chromium just sits on the |
This PR contains:
Breaking Changes?
If yes, please describe the breakage.
Changes the default configuration for the http2 server if
hmris true.Please Describe Your Changes
relates to #183 and #184
I added the
tls-keygendev dependency in order to generate the keys, unfortunately browsers do not seem to support unencrypted http2 at this time, even on localhost. Probably worth some followup work to make the http2 option require some of thehttp2.createSecureServeroptions to avoid any footguns?