[maemo-developers] Wishlist (was:Re: N800 and USB host mode)

[From: Kalle Vahlman]

2007/3/8, Eero Tamminen <eero.tamminen at nokia.com>:
> Hi,
>
> ext Simon Pickering wrote:
> >>> Hitting the Wishlist is almost only a matter of writing down
> >>> a sensible rationale. You can do it.
> >>> (..)
> >>> I will put "USB host mode" in the Wishlist. I plan to
> >>> update the roadmap page on Fridays (if we have stuff to update).
> >
> > Wish #1:
> >
> > I'd like to request PowerVR chipset support (kernel driver, opengl and opensvg
> > implementations).
> >
> > Rationale:
> >
> > The reasons are mainly superficial, however these are valid when you consider
> > that users will probably like flashy games and interfaces, and therefore more
> > devices may be sold (it will also make developers happy to have more cool
> > hardware to play with!).
> >
> > The GUI could take advantage of the snazzy 2/3D features and therefore look
> > nicer as well as being faster, as could 3rd party applications. Things like
> > translucency would be both pretty and useful (e.g. full screen keyboard - thanks
> > Tak). There are many other features and applications that I'm sure people could
> > come up with if they had the option to use it. Game playability (and/or the
> > range of games) would be improved having the 3D acceleration, anti-aliasing,
> > fog, etc., effects.
>
> Err.  Translucency means compositing and keeping the composited items in
> memory.  Due to additional memory accesses needed for this, it would be
> slower (and take more memory) regardless of how "accelerated" it would
> be.

I guess the most appealing reason to have acceleration for graphics[1]
is to help the end result be always available when that slow bus is
ready to push more stuff, and also to enable that stuff to have more
bling.

I haven't confirmed this, but according to the (non-exhaustive)
benchmarking currently it probably should be faster to work with image
surface in Cairo and only push the result to X, since it is
accelerated in neither. This is of course assuming that Cairo's
software implementation is faster to do it, which I don't know. What I
*do* know is that Cairo's image backend gets a nice boost from using
VFP so at least in that case it should beat Xlib surface hands down
(which didn't benefit from VFP that much).

But if the xlib surface would operate with the help of HW acceleration...


[1] Which is not equal to translucency although somehow it always gets
connected to that. I just wonder why..?

-- 
Kalle Vahlman, zuh at iki.fi
Powered by http://movial.fi
Interesting stuff at http://syslog.movial.fi