[Scummvm-devel] WebOS port uses keymapper

[John Willis]

> On Thu, Oct 13, 2011 at 07:39:43AM +0200, Klaus Reimer wrote:
> > On 10/12/2011 08:38 PM, Tarek Soliman wrote:
> > > I just noticed that webos force-enables the keymapper code in
> configure.
> > > I don't think the keymapper code is ready for primetime yet and so I
> > > don't think any release should ship with it.
> >
> > It's the first thing I hear about that the keymapper isn't ready to be
> > used. The webOS port already used it in the 1.3 release and it won't
> > work without it. I have to hardcode some keymappings which are
> > currently configured in the shipped default scummvmrc when I can't use
> > the keymapper any longer. Shall I do this or is it ok if I keep using
> > the keymapper instead?
> 
> It is entirely up to you. I think it is a mistake to ship a release with
the
> keymapper in its current state.

Just to dive in, I can't help but think the use case for WebOS is basic and
may be tested enough to ship for that platform.

We also have the HP TouchPad work that builds on the WebOS port in the patch
queue to consider (I expect it expands on this, not checked mind you) and
the fact it has gone round one full release cycle with it enabled and moving
back to static keymapping may well introduce regressions on the port.

Of course this is just IMHO as another backend hacker not the WebOS guy ;-).
If it is working well for the limited use case..... ;-)

> Keymappings thus far have been done through hardcoding them (see
> Maemo and SamsungTV for some very simple examples but there are many
> others).
>
> There are usages for the keymapper in other ports (e.g. Android) but it is
> not enabled by default and so I assume it doesn't ship with it in the
release
> builds.

Not 100% sure that is the case for all Android mashups. Actually, as it is
Andre I suspect buildbot and release do match exactly for official builds
:).

Regards,

John