- From: Julian Reschke <julian.reschke@gmx.de>
- Date: Sun, 8 Jun 2025 18:16:28 +0200
- To: ietf-http-wg@w3.org
On 08.06.2025 17:08, Julian Reschke wrote: > On 04.11.2024 17:07, Julian Reschke wrote: >> Hi, >> >> I just updated >> >> https://github.com/greenbytes/java-http-1xx-tests >> >> with tests for this scenario. >> >> Of the tested Java clients that support 1xx, only one (OkHttp) currently >> fails (and I expect that to be fixed quickly). >> ... > > I just went back to these tests and noticed that > java.net.http.HttpClient indeed fails this test (many 104s), as it > aborts the request after eight 1xx responses. > > One could even consider this a regression compared to > java.net.HttpURLConnection (as of JDK 20), as that one manages to > consume all 1xx responses. ...but then I was informed that this is only a default, see: https://docs.oracle.com/en/java/javase/24/docs/api/java.net.http/module-summary.html#jdk.httpclient.maxNonFinalResponses Best regards, Julian
Received on Sunday, 8 June 2025 16:16:33 UTC