From 3304041e88c75c40504400c03c3063355b0263ab Mon Sep 17 00:00:00 2001 From: vibesoftwarecoder Date: Thu, 5 Nov 2026 15:48:20 -0600 Subject: [PATCH] Release 0.6.26 Co-Authored-By: Claude Sonnet 5.5 --- docs/release-notes/0.6.26.md | 44 ++++++++++++++++++++++++++++++++++++ version.txt | 2 +- 2 files changed, 45 insertions(+), 1 deletion(-) create mode 100644 docs/release-notes/0.6.26.md diff --git a/docs/release-notes/0.6.26.md b/docs/release-notes/0.6.26.md new file mode 100644 index 0000000..351587a --- /dev/null +++ b/docs/release-notes/0.6.26.md @@ -0,0 +1,44 @@ +**This release adds an option to turn on ApolloVibe's second video stream for seats, and fixes a misleading error in the optional USB driver install.** + +## What changes + +- **A new option, `MultiSeat:SecondVideoStream` (off by default).** When it is `true`, every seat's Apollo is written with `second_video_stream = enabled`, so a MoonlightVibe client can ask for a second picture of a window in that seat. When it is `false`, the config says `disabled`. The seat config is rewritten each time a seat is provisioned, so the setting follows the option after the next provision. Set it in `appsettings.local.json`, then restart the service. +- **A seat can show a window, not a whole display.** A seat is a remote-desktop session, and Windows will not add a virtual monitor to one. If a client asks a seat for a display, it gets one stream. A display works on the console host, which needs no MultiSeat setting. +- **The USB driver signature check now works when Windows PowerShell is started from PowerShell 7** (#141, refs #140). PowerShell 7's module folders came first in the module path, so `Get-AuthenticodeSignature` returned nothing and the check reported `signature status is ''` for a genuine Microsoft-signed file. The check now loads the right module first, and an unreadable signature is reported as unreadable. Nothing is relaxed: a file is still accepted only with a valid signature, the exact signer and the pinned root. +- **The README** covers the second screen, a corrected port table, and a split troubleshooting entry. + +## Who this affects + +- **Everyone who leaves `SecondVideoStream` off**, which is the default, sees no change to their seats. +- **Anyone who runs the optional USB driver step from a PowerShell 7 session.** Before, a genuine driver could be refused with the wrong reason. + +## What the second stream needs + +- An ApolloVibe build from `v2026.6.1-ms8` or newer. MultiSeat's installer fetches the latest release, so a new install has it. +- MoonlightVibe 6.4.0 or newer on the client, with "Show a second window" turned on. + +## How this was checked + +- **The seat option:** seven new unit tests cover the written config line for each setting, and the setting following the option on the next provision. The CI run on `master` passed. +- **On the reference host:** the Gaming seat was re-provisioned with the option on. Its Apollo log read `second_video_stream = enabled`. +- **The USB fix:** reproduced on a real host with PowerShell 7's path first, then fixed. The new test fails when the fix is removed. + +## Not verified + +- **A second window from a real seat** was not tried. A window stream ran in a remote-desktop session like a seat's on a scratch host, and a display stream ran on the console host. +- **The USB fix on a CI runner.** It is not known whether the one timeout in #140 came from this cause. +- **The full test suite** for the USB change was not run locally. It ran in CI. + +## Install + +```powershell +.\scripts\install-service.ps1 -FromZip .\multiseat-windows-x64.zip +``` + +If an isolated Steam is running, the installer stops without changing anything; close Steam first, then run it again. Afterwards, confirm what landed: + +```powershell +& 'C:\Program Files\MultiSeat\MultiSeat.Service.exe' --config +``` + +It should report `0.6.26`. diff --git a/version.txt b/version.txt index 5bfb865..ead87ff 100644 --- a/version.txt +++ b/version.txt @@ -1 +1 @@ -0.6.25 +0.6.26