Understanding queue management during seeding and its different implementation between the Windows client and qBittorrent-nox (WebUI) on Linux

[Dr4w]

I don't fully understand how the seeding queue is managed by qBittorrent when Torrent Queuing is enabled. Specifically, I noticed that we seem to lack control over the queue during the seeding phase. For example, I can't manually choose which torrents should seed and which ones should remain in the queue waiting for their turn. How exactly does qBittorrent handle the queue in this scenario?

On a related note, I observed a difference in behavior between the desktop client on Windows and the Linux client using the WebUI. On the Windows client, I can force specific torrents to seed while keeping the rest in the queue. I achieve this by pausing all torrents, starting the ones I want to seed, and then resuming the rest. This method ensures that the selected torrents continue seeding, while the others remain queued. However, I couldn't replicate this behavior in the WebUI. When I attempt the same steps, the torrents I select are placed back in the queue, and other torrents are randomly (?) chosen to seed instead when everything is started.

[thalieht]

https://www.libtorrent.org/manual-ref.html#seeding-queue
Note that none of the 3 settings mentioned in the end of that section are exposed in qBt so whatever libtorrent has set as default for them is used.

On a related note, I observed a difference in behavior between the desktop client on Windows and the Linux client using the WebUI.

You don't mention versions so one can only assume that different libtorrent versions are used.

[Dr4w]

Thanks for answering my first question.

Regarding the versions, they're both on the latest version, to be more specific:
On Windows: v5.0.2 official client from https://www.qbittorrent.org/download
On linux I'm using a container maintened by linuxserver: v5.0.2 from https://github.com/linuxserver/docker-qbittorrent/releases/tag/5.0.2-r0-ls367