[ag-automation] neuer Mitleser

[Robert Schwebel]

On Mon, Apr 10, 2006 at 12:36:36PM +0200, Jan Kiszka wrote:
> > That was also the case in the good old RTAI days. It is a
> > non-argument.
> 
> No, it wasn't (as long as you do not refer to pre-24-1.x, which was
> before my time).

I've ported several of the patches to later versions in the 24.1.x era
and it was never the patch which was the problem but the fact that the
RTAI people didn't follow the kernel pace. 

> > Problems usually arise when you go to non-mainstream architectures, have
> > non-mainstream environments and non-mainstream features. A long term
> 
> See blackfin port or IA64 support.

For example. The thing is, when internal APIs in the kernel are changed
all in-kernel stuff is changed as well. All external code has to follow
actively. 

> Communities are moving, as well as their interests, and there is not
> only one. Xenomai, e.g., strongly benefits from "the" community, check
> the lists and and related projects. And it will surely benefit from a
> kernel and glibc RT-community as well in the future.

Ack. 

> > So "rt_open()" would be "open()" when using the POSIX skin?
> 
> Yes.

Ok, then it would be a good topic for the meeting to find out how to
design the driver infrastructure for industrial I/Os to be able to work
with both realtime variants. That would indeed be an interesting aim,
because it influences lots of decisions.

Thomas, do you have some material available regarding your IIO
framework? That would be a good base for further discussion. 

Robert
-- 
 Dipl.-Ing. Robert Schwebel | http://www.pengutronix.de
 Pengutronix - Linux Solutions for Science and Industry
   Handelsregister:  Amtsgericht Hildesheim, HRA 2686
     Hannoversche Str. 2, 31134 Hildesheim, Germany
   Phone: +49-5121-206917-0 |  Fax: +49-5121-206917-9