I prefer using a mode, it seems more idiomatic. Shall this use
cl-defgeneric in sleep-wake-mode that each system can specialize?

On Thu, Feb 6, 2025 at 8:02 AM Andrew Cohen <acohen@ust.hk> wrote:

> >>>>> "SM" == Ship Mints <shipmints@gmail.com> writes:
>
>     SM> The API on macOS would expose a pair of C functions to lisp, one
>     SM> would inhibit sleep, perhaps returning an opaque "cookie", the
>     SM> second to revoke the inhibition. The same API can be used to
>     SM> prevent macOS from ever sleeping, if the user wanted that.
>
> Great!
>
> So the idea now is:
>
> Turning on sleep-wake-mode:
>
> 0.  If on a system with D-Bus and logind, we register a signal for
> PrepareForSleep with a handler that generates a synthetic sleep-event on
> state change using Michael's recent insert-special-event.
>
> 1. Install 'sleep-event-handler on the special-event-map for
> [sleep-event].
>
> 'sleep-event-handler will inhibit sleep (with whatever method is
> appropriate for the system); run the function sleep-handler-function
> which is a defcustom defaulting to running the hooks; release the
> inhibition.
>
> 2. disabling sleep-wake-mode removes 'sleep-event-handler from the
> keymap and replaces it with 'ignore.
>
> (alternatively we could ask each system to expose a lisp function or
> variable to turn on and off the generation of sleep-events)
>
> Best,
> Andy
>
>
> --
> Andrew Cohen
>