After the Gatemiga paper switch to Linux, QNX the company predicated its moves to work on a user desktop on phase5 partnership. When phase5 no longer showed any ability to pursue its part of the partnership, QNX still had some hope that it could leverage its plans toward a self-hosted DEVELOPMENT desktop to dovetail with a sustained climb toward multi-hardware-platform user desktop amenities, including improvements on Amiga-specific concepts such as datatyping, and process managers providing compatibility layer/APIs, and some more efforts to widen the application base beyond *nux/nix ports.
But when less and less Amiga community interest was shown because a lot of the Amiga-conversant developer base they hoped would participate - with a need for a deep partner(s) with phase5-level skills - was instead being drawn toward Amiga DE development, they began to prioritize differently, leveraging the demand they enjoyed in telecom, aerospace, mobileGT, and industrial and embedded even more than before.
Because of breakdown in negotiations between Amino (now the new Amiga Inc) and QNX (QNX could see no realistic business plan from Amino and did not like the Amino approach toward relationship-building), Tao was hastily chosen to keep from dieing stillborne - either because (1) Amiga Inc didn't do its homework in looking for an alternative that would do what they claimed they wanted to do at Gateway2000, or (2) they had no real intentions to deal with the desktop and the home gateway concepts, instead just wanting to string the community along while they made further inroads with developers.
This really sapped the development pool and momentum QNX was interested in that would have complimented the *nixy types pretty much guaranteed by POSIX compliance, and the embedded developers who were already working with the older QNX (pre-6.x) codebase. After alpha and beta testing and some early apps ports it became obvious to Phoenix that some of us had to look for major finance and talent to build on QNX for our needs. Indeed, most of us preferred QNX for modern desktop potential and for embedded spinoffs.
Fortunately, after dealing with sharks of various tooth size and not wanting to sell all our work for loss of control the VC sharks insisted on, we saw that Genesi was forming and had a strong interest in multi-OS work, and in both desktop and embedded work. Just in time, from my perspective!
Today its pretty obvious that Tao's multiplatform approach was designed for other uses entirely, and it seems a shame that QNX's multiplatform approach wasn't finally to get more desktop emphasis. But QNX itself has done very well in expanding its base and its potential, which only seems to intersect the desktop for the average user slightly.
Phoenix still has people doing QNX stuff, and Genesi is involved in making sure QNX is ported and that a business relationship with QNX is thereby developed. QNX is waiting for Peg2 after having ported to Peg1. When Peg2 porting and testing is done (should go fairly quickly given the speed of getting it happening with Peg1) it will be time to seriously consider product using the two.