Better touchscreen support

Registered by Martin Pitt on 2009-11-04

Better touchscreen support

Blueprint information

Martin Pitt
Bryce Harrington
Needs approval
Series goal:
Not started
Milestone target:
Completed by
Martin Pitt on 2012-05-15

Related branches



Work items:
[albertomilone] Identify one or more target touchscreen hardware platforms to focus on: TODO
[albertomilone] Purchase and expense one touchscreen device: TODO
[pitti] Migrate all touchscreen hal rules to udev: TODO
Identify common touchscreen devices not currently supported by -evdev: TODO
Implement support for rotating screen and touch interface together (Bug 217182): TODO
Investigate existing patches to Metacity or other window managers for touchscreen support: TODO
Implement display rotation using multi-finger drag gesture: TODO
Implement zoom using a "pinch" gesture: TODO
Implement larger gnome panel: TODO
Scale up icons to a more finger friendly size: TODO
Implement fatter scroll bars: TODO
Implement larger window resize handles (or force all apps to be full screen or fixed pos/size): TODO
Evaluate work for smart identification of "what you meant to hit", ala iphone: TODO
Investigate any existing work done to morph CellWriter into a finger-driven keyboard interface: TODO
Investigate any existing work with other onscreen keyboards: TODO
File a MIR for CellWriter: TODO
Review CellWriter bug reports and plan a triaging/upstreaming program: TODO
Enlarge keyboard keys to a more finger-friendly size: TODO
Implement patch that integrates cellwriter into Gtk so can be easily launched when entering data into any text widget: TODO
Enable easy triggering of CellWriter via iBus or other appropriate protocol: TODO
[bryceharrington] Document test cases on X wiki: INPROGRESS
Create test cases for appropriate applications: TODO

bryce 2009-11-24: Drafted blueprint based on UDS discussion.

bryce 2009-11-25: I'm taking myself off as assignee because most of the work items are UI tasks

pitti 2009-11-27: The entire set of actions is obviously too big for lucid, and some (like larger gnome panel) can't even be done by default on all devices (that would need a dynamic "do we have a touch screen" detection), but this is a nice target of opportunity. I propose we start on the "right" way for the kernel drivers, even if it just works for very few devices (which would also be the case for the quick'n'dirty hack anyway).


Work Items