Repository navigation
XHTTP mode choice with REALITY in Nov 2026, and empty-SNI detection (ref #7280) #7418
Unanswered
a8813090-bot
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Two practical questions from a user in Russia, November 2026. Both are about current recommended practice rather than bugs.
1. XHTTP mode selection with REALITY.
Community write-ups currently recommend
stream-one,packet-upandautorespectively, each presented as the correct choice for REALITY over TCP/443. One article additionally states that when upload and download are split across two addresses viadownloadSettings, onlyautoworks andstream-oneis guaranteed not to.Is there current guidance on which mode to prefer with REALITY? And does the recommendation differ between mobile networks with aggressive filtering and ordinary home broadband? I am happy to test modes from this vantage point and report back if that is useful.
2. REALITY with empty SNI.
#7280 reports that Aparecium still flags connections with an empty SNI on v26.10.21, because post-handshake records are not imitated in that case, while the non-empty path was addressed in v25.6.8.
serverNamethe recommended mitigation?Thanks for the work on this project.
All reactions