John Nagle wrote:
This looks like the beginning of the end of QNX as
a usable real-time operating system.
Harman International does nothing but audio products.
“If it plays, records, or produces music, chances are it’s
a Harman product”. They don’t do industrial control systems.
They don’t do core automotive electronics. They don’t
do “it has to work” stuff. They’re expanding into car
video players and navigation systems, but those are
all basically accessory items.
Hm … sound studios and broadcasting systems are accessory items??
No … these are complex systems! Your view is a little bit narrowed.
The analog technology for audio and video devices will be dead in few
years and will be replaced be intelligent digital devices for e.g.
digital radio and television systems.
This is key. If QNX had been sold to a company that
built core automotive systems, like braking, steering,
and engine control, that would be one thing.
I don’t believe that QNX is applicable for that low level
(still 16bit) microprocessor applications.
But Harman
only does non-core entertainment-type devices. If your
CD player crashes, it’s not a big deal. This implies
a huge change of focus.
Show me one low level motor control system working with QNX …
Software is more and more the CORE of complex audio/video systems.
An employee from Becker told me some years ago that their car audio
hardware could be considered as a ‘dongel’ for their embedded
software 
That’s the reality and the reason why the control over this core
software technology is realy vital for Harman.
QNX users involved in high-reliability systems that
control real hardware need to be looking for an
exit strategy. Big QNX users need to be talking about
code escrow and contractual protections. Hard real time
control is just too far from Harman’s core business areas.
Digital broadcasting systems have STRONG requierements for hard
realtime! Just an example …
It’s a sad ending for a great technology.
No … I think it’s a great chance for QSSL.
Regards
Armin
John Nagle