maritz

Okay, now the tests passed. I'm hoping that these changes actually fixed it. Can't guarantee it though, since I don't truly understand what caused it in the first place. Anything...

@dougwilson Could you manually trigger the travis checks for this PR a couple of times to see if it still fails occasionally? Not a great thing to have to do,...

Sorry to pester you, but is there any update on this?

There were changes in node v9.8.0 that have the potential to maybe fix this... That's about all I got for this right now. :-(

I'm actually occasionally seeing this in the one app where I use this PR. I have no indication yet what causes it and it always takes the entire process down....

I added a fix for a change in node 10.4 onward that caused test timeouts every time. This allowed me to add node v10.11 to the travis.yml targets, providing one...

Sadly this is still happening, had to revert to the proxy doing compression again. ``` Error [ERR_HTTP2_INVALID_STREAM]: The stream has been destroyed, at Http2ServerResponse.write (internal/http2/compat.js:599:19), at Gzip.onStreamData (/usr/src/app/node_modules/@maritz/compression/index.js:206:20), at Gzip.emit...

Just to clarify: The code changes in this PR should be completely fine - they just change _implicitHeader to writeHead. There is just some problem with how compression works over...

If you can figure out what is causing the seemingly random failures, that would be great. I don't have the time and honestly probably lack understanding of how compression and...

> do you have any hints on what your app does that might be provoking those failures? Not really, it was just failing after X amount of time working fine....