Apache HttpComponents

I am the original requestor of this feature, more than 10 years ago. Normally I would just request that olegk should state what he means by "better mechanisms" but he can't, because there aren't any. A TCP connection can go stale for a variety of Read more

abernal I think implementation of this config policy was a mistake. The feature was requested over ten years ago and since then better mechanisms of dealing with stale connections and out of sequence responses have been created. This config parameter Read more

michael-o commented on PR #615: URL: https://github.com/apache/httpcomponents-client/pull/615#issuecomment-3418550542 > > > @michael-o Any chance this feature could make it into 5.6? > > > > > > Do you have a timeframe for 5.6? > Read more

ok2c commented on PR #615: URL: https://github.com/apache/httpcomponents-client/pull/615#issuecomment-3418526830 > > @michael-o Any chance this feature could make it into 5.6? > > Do you have a timeframe for 5.6? @michael-o I would Read more

RequestUpgrade interceptor tries to upgrade the connection when using proxies

abernal This has nothing to do with RFC 2817. This is pure RFC 9110, the Canon of the HTTP specification at the moment. Those proxies should improve their conformance and not us adding more config knobs. This upgrade feature is actually useful Read more

This is what RFC 9110 states. ``` 7.8. Upgrade The "Upgrade" header field is intended to provide a simple mechanism for transitioning from HTTP/1.1 to some other protocol on the same connection. A client MAY send a list of protocol names in the Read more

olegk  Unfortunately, as far as I can see, the proxy details are not being logged. 2025-10-14 20:16:40,135 DEBUG [main][org.apache.hc.client5.http.headers] http-outgoing-0 >> Upgrade: TLS/1.2 2025-10-14 20:16:40,135 DEBUG Read more

Virtual Thread (JDK 21) support for classic CloseableHttpClient — disabled by default

ok2c commented on PR #717: URL: https://github.com/apache/httpcomponents-client/pull/717#issuecomment-3393586089 @arturobernalg Likewise, I do not really understand the benefits of this change especially for the classic transport where the users Read more