Mer/Packages/powerlaunch

(draft)
({{Mer}})
 
(2 intermediate revisions not shown)
Line 1: Line 1:
 +
{{Mer}}
 +
Powerlaunch Re-implementation
Powerlaunch Re-implementation
Line 23: Line 25:
The daemon doesn't have all the complex config parsing stuff resident all the time; also need to avoid inventing a scripting language on top that wires together the stuff.
The daemon doesn't have all the complex config parsing stuff resident all the time; also need to avoid inventing a scripting language on top that wires together the stuff.
 +
 +
==Implementation Proposal==
 +
 +
Initial implementation of the outline design using a python launcher and even plugins then review the use cases and see how it impacts memory usage. If we find we can keep it simple then we port it to vala.

Latest revision as of 11:12, 22 September 2010

Image:Mer infobox icon.png
Mer is back : Please visit http://www.merproject.org to learn more about this news.


Powerlaunch Re-implementation

suggestion: very simple vala daemon which is essentially a object/plugin loader, providing some degree of conflict-prevention, run a glib main loop and let the behaviour implement themselves as they want

[edit] Use Cases

Needed

Eg Dim screen on Idle when the device is 'locked' the key handling connections aren't made *apart* from long press on the power

conditional and grouped instructions

[edit] Proposed Design

Vala based daemon which is driven by a validated machine parseable config (cf http://live.gnome.org/Vala/DBusClientSamples)

Glib plugins to this daemon present UI-facing methods.

Python config parser to generate validated machine parseable config from a human readable (or gui managed) description. Essentially joining signals and slots from the plugins; possibly with conditionals.


The daemon doesn't have all the complex config parsing stuff resident all the time; also need to avoid inventing a scripting language on top that wires together the stuff.

[edit] Implementation Proposal

Initial implementation of the outline design using a python launcher and even plugins then review the use cases and see how it impacts memory usage. If we find we can keep it simple then we port it to vala.