Bryan Childs wrote:
Ladies,
For those of you not currently paying attention a brief summary :
At time of writing we have three current applications for Google Summer
of Code projects, one of which is a very nicely written one for further
developing Rockbox as an Application. During review of this, it's been
pointed out (by the ever insightful linuxstb aka Dave Chapman) that
whilst we've had this on our list of "Stuff we'd like to see happen" for
a longish time now, there's no real clear idea of what it is we want
exactly. The current feeling amongst those that have discussed this
thorny issue is that we should perhaps try and have some discussion
around that before GSoC really gets going, so that we don't end up
pulling the carpet out from under our student (should he be accepted).
My personal take on this is perhaps not hugely detailed, and I've been
willing to let whoever took the first crack it really have free reign -
working on the assumption that "some code is better than no code".
However others feel differently it seems, so now is your chance to speak
up.
The choices at a broad level appear to be :
1) Make RaaA use all of the existing Rockbox UI code, and just talk to
that code with whatever interface is availble on the device running the
application.
2) Make RaaA not use any existing Rockbox UI code, and implement some
sort of framework for controlling rockbox via any available native UI.
This would probably need to take into account threading, button
handling, sound and UI widgets.
Anyone got any comments ?
Bryan
I would like to suggest that a common framework across environments
would be wise, e.g., something that is available for Windows, Linux,
BSD, OS X, etc. Qt? With a layer in there to fit in Paul Louden's
idea, so that there is minimal wear and tear on the RockBox code.
Mark Allums
