Remove NPN support from OkHttp #1201
Conversation
Work in progress. I see a new failure from a certificate pinning test but I don't know why. Help solving this would be appreciated. I also see failures before & after: 1) A failure from DisconnectTest (square#1197 may fix that). 2) A failure from CallTest.doesNotFollow21Redirects_Async (timeout) The maven changes should be treated with the contempt they deserve. I mostly removed things I didn't understand and stroked maven until it stopped squealing. The benchmarks / okcurl changes are a particular mystery. Tried with arbitrary versions of openjdk7 and openjdk8 I had lying around. Behavior was the same across both (i.e. the same failures).
|
The certificate pinning test is flaky. I need to fix that. |
| @@ -97,14 +97,15 @@ | |||
| </build> | |||
| <profiles> | |||
| <profile> | |||
| <id>npn-when-jdk7</id> | |||
| <id>alpn-when-jdk7</id> | |||
There was a problem hiding this comment.
There's no such thing as ALPN support for JDK7. Dropping NPN means dropping SPDY and HTTP/2 from JDK7 completely. I'm okay with that.
There was a problem hiding this comment.
Are you sure?
Based on:
https://webtide.com/jetty-9-2-0-released/
"ALPN is the successor to NPN (the Next Protocol Negotiation) and is implemented for both JDK 7 and JDK 8. ALPN will be used by SPDY and HTTP 2.0 to negotiate the application protocol of a TLS connection."
Various SPDY tests broke on OpenJDK 7 without this, and passed with it.
There was a problem hiding this comment.
Wow, news to me. That's excellent.
|
So, some sites only support NPN and may not switch SPDY over to ALPN. Also, I don't think twitter responds to ALPN at the moment, so integrations like twitter4j will break. @jpinner am I right on this? |
|
OpenSSL does not currently support ALPN (it's slated for 1.0.2 currently in beta). Twitter (and I suspect other sites running OpenSSL) will not be able to support ALPN until 1.0.2 lands. |
Work in progress.
I see a new failure from a certificate pinning test but I don't
know why. Help solving this would be appreciated.
I also see failures before & after:
(Apply fix for devices with large socket buffers #1197 may fix that).
The maven changes should be treated with the contempt they deserve.
I mostly removed things I didn't understand and stroked maven until
it stopped squealing. The benchmarks / okcurl changes are a
particular mystery.
Tried with arbitrary versions of openjdk7 and openjdk8 I had
lying around. Behavior was the same across both (i.e. the same
failures).