Repository navigation
Replies: 1 comment 1 reply
|
Thanks for running this and writing it up so carefully. I've never been able to try the iOS bind myself, so this fills a gap I couldn't test. I've updated the docs in #57. The "not tried live" notes now point here, as a third-party report I haven't reproduced. The own-device section lists what you saw: the bind, the Your log covers a bit more than your caveat says. The bind's first step is the same iOS guest One thing I'm still curious about, if you still have the log: did that refresh return a new |
Uh oh!
There was an error while loading. Please reload this page.
Summary
auth.mdxnotes, for the device-of-its-own flow:I tried it on 2026-10-28, and it works end to end: a fresh, Go-generated iOS device binds to a real account by email verification and logs in, loading the account's full game state. So the "iOS untested" gap is now closed on the iOS side too.
What was run
Bind (one run, Android identity replaced by iOS throughout via
-cs-ios):Then the plain login test:
Results (log, redacted)
Bind:
Login with that config:
Observations relevant to the repo
shumeiBoxIdempty is accepted on iOS, same as the established-account case. The iOS ShuMei scheme does not need to be reversed for this path —-cs-shumei-genis correctly ignored in iOS mode and is not required.tagate is the same on iOS: the accepted login carries the filled iOS analytics blob with the platform markers (#os=iOS,#bundle_id=com.lastwar.ios,lw_platform=AppStore, the build,lw_zone), andlw_device_id/lw_shumei_id/lw_airKeystayed blank.ta={}was not retried here, but the filled blob worked first try.appVersion=1.0.344/versionCode=786(the iOS build-cs-iosclaims), and the base-zone login under that identity is accepted.opt=refreshon the iOS own-devicertworks too: a follow-up GSLopt=refreshwith the config's refresh token returned a fresh access token (code 0), which answers the related open note that an iOS refresh "hasn't been tried live" — on an own-device pair (which has anrt), it does.Caveats
All reactions