Say where persistence stops: sessions do not survive logout #21
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Context
Sessions outlive their windows because the agent owns the pty. They do not
outlive a logout: the agent is started on demand, without a systemd unit and
without lingering, so the user's session scope is torn down and every session goes
with it. This is documented in
docs/spec/session-agent.md:It is in the README too. It is not anywhere the user can see while using the
program.
The problem
"Persistent" is a per-session setting with a switch in the session editor. A
reasonable reading of that switch is "this session keeps running", and it does —
until logout, when it does not. The place to say so is next to the switch, not in
a file on a git server.
Work
Small, and mostly wording:
crates/terminalko/src/editor.rs/ui/session_editor.blpgains a subtitlesaying what it does and does not survive: window closed yes, logout no;
ui/preferences.blp) already explainsthat sessions keep running when a window closes. One more sentence about logout
belongs there, with the
loginctl enable-lingerhint;loginctl enable-lingerneeds polkit and is a system change, so a button that does it silently is wrong;
a copyable command is honest and enough. If issue #18 (socket activation) lands,
revisit — with a unit installed, lingering actually makes sessions survive
logout, and then the wording changes from "does not" to "unless you enable
lingering".
Acceptance criteria
a user can act on.
docs/spec/preferences.mdanddocs/spec/session-agent.mdstay consistent withthe new wording.
Notes
losing work they thought was safe.