Two messages:

**************************************************

This is about packages which contain graphics devices: I identified

Cairo GDD JavaGD RSVGTipsDevice RSvgDevice cairoDevice canvas devEMF rscproxy tikzDevice

I have noticed that some graphics devices are not setting some of the hooks introduced in R 2.11.0:  in current R (it is different in R-devel) there will be a segfault if the corresponding R-level functions are called.

So please ensure that you have at least dummy entry points dd->cap and dd->raster set (the first should return R_NilValue).

Secondly, there is the opportunity to add some new facilities in R-devel, to be released as R 2.14.0 on October 31.  These are the components

devHoldFlush()
haveTransparency
haveTransparentBg
haveCapture
haveRaster
haveLocator

documented in src/include/GraphicsDevice.h.  (If you include them, you should condiitionalize on

#if R_GE_version >= 9
#endif

.)  Screen devices (cairo*, JavaGD ...) will benefit from implementing dev.hold/dev.flush (see their help page), and devices that implement rasterImage() should set haveRaster accordingly: e.g. in the near future image() will use rasters where available.

The remaining components support the new dev.capabilities() function. More may be added in future, but if the components are unset they are initialized to 0, which is interpreted as NA by dev.capabilities().

Brian Ripley

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595


******************************************************************************************

It seems the path() component (introduced in R 2.12.0) is missing in packages

Cairo GDD JavaGD RSVGTipsDevice RSvgDevice cairoDevice canvas rscproxy

There is no protection in the graphics engine in 2.12.x and 2.13.x against calling that component.

On Fri, 2 Sep 2011, Prof Brian Ripley wrote:

> This is about packages which contain graphics devices: I identified
>
> Cairo GDD JavaGD RSVGTipsDevice RSvgDevice cairoDevice canvas devEMF rscproxy tikzDevice
>
> I have noticed that some graphics devices are not setting some of the hooks introduced in R 2.11.0:  in current R (it is different in R-devel) there will be a segfault if the corresponding R-level functions are called.
>
> So please ensure that you have at least dummy entry points dd->cap and dd->raster set (the first should return R_NilValue).

AFAICS only cairoDevice, devEMF and tikzDevice have dd->cap: those and Cairo have dd->raster.

> Secondly, there is the opportunity to add some new facilities in R-devel, to be released as R 2.14.0 on October 31.  These are the components
>
> devHoldFlush()
> haveTransparency
> haveTransparentBg
> haveCapture
> haveRaster
> haveLocator
>
> documented in src/include/GraphicsDevice.h.  (If you include them, you should condiitionalize on
>
> #if R_GE_version >= 9
> #endif
>
> .)  Screen devices (cairo*, JavaGD ...) will benefit from implementing dev.hold/dev.flush (see their help page), and devices that implement rasterImage() should set haveRaster accordingly: e.g. in the near future image() will use rasters where available.
>
> The remaining components support the new dev.capabilities() function. More may be added in future, but if the components are unset they are initialized to 0, which is interpreted as NA by dev.capabilities().
>
> Brian Ripley
>
> -- 
> Brian D. Ripley,                  ripley@stats.ox.ac.uk
> Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
> University of Oxford,             Tel:  +44 1865 272861 (self)
> 1 South Parks Road,                     +44 1865 272866 (PA)
> Oxford OX1 3TG, UK                Fax:  +44 1865 272595
>

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595
