in case you are interested in hiring or buying please contact me. I am working on an industrial embedded real time ARM Cortex A8 platform, derivated from BBB running QNX. I am in cooperation with another small company that is doing the hardware derivate. Software is on my part.

Thorsten Gonschior
www.spectral process.com

NEWS: - Show stopper -

After clearing up some technical aspects I ran into a major show stopper. Price of QNX Develoment environment is about 10.000,-eur. Even if I would be willing to invest this kind of money, my tarket was to establish a development platform for engineering offices. I know for sure that most of small engineering offices or soltary engineer are not willing or able to invest this for smaller of projects.

I start the attempt to convince QNX that there is a large community out there willing to work with qnx OS for a huge variety of projects and applications, but not willing to passt this kind of entry barrier.

I would appreciate comments on my assumption.


Well that was my point in my previous comment. Whereas I think QNX is a really great RTOS hopefully “brigded” to opensource -whereas being closed source- in recent years (yes recent compared to >20 years of existence) , from the start I’ve believed in the µKernel architecture as the best one. Anyway, prices and -royalties- allways stopped me, as for µC/OS as well.
Honestly, I really like to use QNX on BBB but strongly believe that if it’s not on Raspberry it won’t be ported to BBB neither.

quote: “Seems not so obvious. So QNX has an opensource part but for BBB it implies these steps …”

hm, ok sry I must have missed your point :slight_smile:

I agree that QNX is probably the best choice for a robust and well designet real time embedded system. In this I do not bother the OS-price or royalty which probably is scalable.

Texas Instruments learned this 10 or 15 Years ago when the Code Composer Studio was very expensive and a simple JTAG Pod cost 3.000,-. Even Microsoft pushes the development platforms for free into the market, because all of them understand that the major market is selling the operating system (in case of Microsoft) or selling the hardware (in case of TI) and not selling the development solutions. See XILINX, development studio incuding simulator for free.

This is what makes people work with it. Expecting 10.000,- as entry cost is the best guaranty to avoid thousands of developers working with QNX and thereby hundreds of thousands licences of the OS.

I dont understand this philosophy …

20 Years of experience as embedded developer showed me on thing. 1.000,- eur or $ ist the magic barrier. this is what people invest as an attempt for some half cooked idea thy have. Max be that 95 % of those ideas will never reach any state of economic asset, but once in a while it will and it will do with power. suppressing this kind of innovative forces is hurting the provider itself.

I am 50% convinced today that I will build my platform on Linux RT.

keep you posted …

well ok, but I’m not sure to understand all of it :

“I am 50% convinced today that I will build my platform on Linux RT”

what are the other 50% stands for ? what will change it for a complete decision, what are the requirements above price ?
For the “freeing” part I’d say that QNX did it differently than the other (MS, Ti, and even Arm ) with their licensing policy. Anyway, this doesn’t helped so much indeed: hence my particular focus on Xenomai for BBB instead.

the 50% means that I still hope for QNX to make a better price for the development system (much better). The local Distributor is making an appeal to qnx in this matter, especially to reduce costs for the endcustomer developer. My reality proven me tells me to accept the situation as it is and find something payable.

the other 50% hopes for RT patch on linux ist better than my experience tells me or xenomai is less complex than I currently think. My research on this is ongoing. I am currently reading reports, experience papers and stuff with suspicious conscience.

QNS would have been a nice choice. probably it just takes some time for me to convince myself on a chemical basis. lets say cheat my guts feelings accept the facts deal with it … :wink:

RIM has lots of money they need to keep coming in and cutting prices means less income so don’t hold your breath or you’ll never get the project done but you also won’t be around to worry about it neither. :wink:

FYI, they paid around $200M for QNX and their other businesses are not doing well…


understood :slight_smile:

but think of this… 5.000 development licenses @ 10.000,-€ → 50.000.000
plus 5.000 projects decoupling 1000 OS licenses @ 100€ each → 500.000.00,

ok I am just guessing

500.000 development licenses for free → 0,-
plus 20.000 projects decoupling 1000 OS licenses @ 100,- each →,-

play with the numbers all you want, but the principle stays the same

development entry costs supress the multiplicator of having many active products in the market that sell OS licences :wink:

in addition to this, many products in the market provide many interested pepople building many new applications that brings you back to my point … :smiley:

again, TI did it, Microsoft did it, Xilinx did it, and I guess a lot of others too.

yea who am I to question … I am currently deeply interested in xenomai … crazy stuff,

" He’s a lumberjack, and he’s ok. He sleeps all night and works all day."

Seems you went into the same path then : left QNX for Xenomai.
Btw, putting this in the previous post this decreases the potential additional licenses from the 500.000 you were guessing about.
I talk about people that won’t buy license and almost won’t put any energy to get QNX more involved, as instance in BBB here. These are many loss.

QNX also have some interesting
For MicroKernel architecture, You can make GLES2 app run within 2 seconds from power on
On current Ti BSP ,whatever GLSDK / EZSDK/ DVSDK
I dont know if anyone get GLES app can run within 2 second