Comment 42 for bug 638307

> With 1.57, I've already dropped that reliance on pm-utils ... depending on what
> infrastructure is available (pm-utils/acpid/kernel events), we trigger
> lmt (with proper locking and not triggering the same again).

Then what is the conflict here?

Pm-utils ships a suspend infrastructure and scripts (directory tree for hooks), and l-m-t hooks into them providing sensible human-readable configuration options.

Pm-utils also includes a bare-bone laptop_mode script (like ubuntu-laptop and acpi-support did before, and have dropped again before). Did the pm-util laptop_mode script stop to check whether the laptop-mode-tools is installed?

How can there be a confict between the pm-utils package that provides the hooking facilities and any package that installs hooks into it?

Shouldn't this be avoidable by the ordering of the hooks (i.e. by providing an early 00-build-in-pm-utils-laptop_mode that gets overridden by later hooks, or no-ops). Or alternatively, by l-m-t installing its hooks under /etc/ overriding the hooks under /lib?