Linked by Thom Holwerda on Wed 12th Mar 2014 23:05 UTC, submitted by poundsmack
QNX

QNX 6.6 has been released. There are quite a lot of changes, but - sady - I doubt many of us work with QNX itself. It's quite popular in the embedded world, and, of course forms the foundation for BB10 - which has not exactly been a stellar success. Anywho, there's a pretty big change in 6.6:

The new Screen Graphics Subsystem replaces the Photon microGUI, including PhAB, Phindows, and QNX Neutrino Advanced Graphics. Usually referred to simply as "Screen", the Screen Graphics Subsystem allows off-screen rendering and can composite graphics from different rendering technologies, including HTML5, Elektrobit GUIDE, Crank Storyboard, Qt, and native (e.g., OpenGL ES) code.

Photon has been such a core part of QNX' identity for me that it's kind of weird to see it go.

Order by: Score:
QNX dominates the in-vehicle world
by tanishaj on Thu 13th Mar 2014 00:48 UTC
tanishaj
Member since:
2010-12-22

Implicit in "quite popular in the embedded world" is the fact that QNX dominates in-vehicle systems. QNX customers include Audi, BMW, Chrysler, General Motors, Honda, Mercedes and Toyota to name a few. There are rumours that Ford will move away from Microsoft to QNX soon.

In fact, Apple's new CarPlay system runs on QNX. So, QNX is doing some pretty exciting things.
http://www.zdnet.com/blackberrys-qnx-why-its-so-valuable-to-apple-g...

It is interesting that Thom felt the need to throw in that BB10 "has not exactly been a stellar success". That may be true but BlackBerry still has 3x the market share of Windows Phone (which Thom likes).
http://www.netmarketshare.com/mobile-market-share

BlackBerry runs Android apps today and has sold millions of units. It remains to be seen how easily Jolla manages the same but Thom is already VERY excited about them to the tune of several articles.

We all have our favourites. I really like my BlackBerry Z30 (switched from an iPhone) so perhaps I am just sensitive.

Reply Score: 5

ingraham Member since:
2006-05-20

It is interesting that Thom felt the need to throw in that BB10 "has not exactly been a stellar success". That may be true but BlackBerry still has 3x the market share of Windows Phone (which Thom likes).


Well, BlackBerry "classic" and BB10 don't seem to be split out in that chart, so we really don't know how well BB10 doing. But "not exactly a stellar success" is a pretty soft way of saying "scrabbling over the 11% of the market left over from Android and Apple." Also, Thom may LIKE Windows Phone, but he has never claimed that it was a stellar success. Likewise his praise of Jolla has included statements that point out the difficulty Sailfish will have in the market.

You want an example of trolling BB users? Try this: BlackBerry will cease to exist before the end of 2014. Due to it's reliance on QNX -- which will survive -- BB10 will never be able to go open source, so it will die more thoroughly than WebOS, PalmOS, and BeOS. (Hey, didn't Palm take BeOS down with them?)

Reply Score: 6

BlueofRainbow Member since:
2009-01-06

It is doubtful that BB10 will ever be open sourced as it runs on a specific set of devices - being based on QNX is only an hindrance as the functionality could be implemented over a Linux Kernel (in a similar fashion to Android for this matter).

Reply Score: 2

tanishaj Member since:
2010-12-22

You want an example of trolling BB users? Try this: BlackBerry will cease to exist before the end of 2014. Due to it's reliance on QNX -- which will survive -- BB10 will never be able to go open source, so it will die more thoroughly than WebOS, PalmOS, and BeOS. (Hey, didn't Palm take BeOS down with them?)


Interesting perspective. I long ago predicted that Android would dominate iOS because it was open. However, I am not sure that Open Source has mattered as much as the open ecosystem (like Windows before it). As a Linux user since 1992 that cannot wait to play with Ubuntu mobile, I am starting to doubt that Open Source matters much in the mobile market. Due to the reliance on Google services, Android is less Open Source all the time and nobody seems to much care.

BlackBerry is not open enough to be the dominant consumer phone platform if that is what you are shooting at. It remains to be seen if there is an "enterprise" segment in devices where BB10 could be the UNIX of mobile. It depends how far BYOD goes.

Saying that BlackBerry the company will cease to exist this year is pretty bold. The quality of their current management alone argues against it. I think Chen will spend a few years building before he sells BlackBerry.

Counted in WatsApp bucks, the current market price for BlackBerry might be worth paying just to get BBM (messenger). BES still dominates the enterprise device management space. BES12 will be by far the most capable cross-platform solution and that business is still growing. Saying that BlackBerry will migrate to becoming a pure software company is a much more arguable position.

Personally, I think the chances of an all software BlackBerry are only a little better than an all devices Microsoft.

Devices is the weakest part of the BlackBerry business. The consumer/retail side even more so. That said, successful niches (like Apple) are still possible. Regardless of profit potential, devices are strategic for BlackBerry on the enterprise side and a real differentiator against software-only competitors. With Foxconn making the devices, BlackBerry does not have a lot of device risk or overhead.

There is of course also the wild card of the tens of millions of current BlackBerry users worldwide. If BB10 starts to get traction, the platform could still be successful. BB10 was only released a year ago after all and is just starting to find it's legs.

In short, I do not find your prediction very credible. Then again, I have been wrong on the Internet before and your opinion is not an unpopular one.

PS. Why would BB10 would die worse than PalmOS or BeOS? Those are also closed source. BeOS was indeed sold into obscurity (to Access Co.) as part of PalmSource. BeOS may live again as Haiku but that goes against your argument as people could certainly produce an Open Source clone of BB10 as well (which would truly surprise me). WebOS never enjoyed any commercial success but it is still out there somewhere thanks to Open Source as you say.

PPS. BlackBerry could certainly Open Source BB10 (including QNX) if they wanted to. They could also Open Source everything but QNX and it could be ported to a Linux kernel. That may not be too hard given that both QNX and Linux grew out of the POSIX ecosystem. I do not see either of these things happening though.

Reply Score: 2

majipoor Member since:
2009-01-22

In fact, Apple's new CarPlay system runs on QNX. So, QNX is doing some pretty exciting things.
http://www.zdnet.com/blackberrys-qnx-why-its-so-valuable-to-apple-g...


No: CarPlay runs on the iDevice iOS 7.

With CarPlay, the iDevice streams its interface to the embedded system which is then able to stream back the user input (touch, button, press on resistive display) using Apple CarPlay protocol (embedded into a H.264 stream apparently).

Currently, the embedded system presented by Apple and Ferrari/Volvo/Mercedes does run on QNX indeed, but saying that CarPlay runs on QNX is like saying the AppleTV runs on the TV "OS".

I guess Apple may have implemented some libraries or some runtime running on the embedded QNX system to help car manufacturers and because Apple want to control everything, but the piece of code which is running here is only able to display a H.264 stream and send inputs to the connected iDevice.

But I know it is quite pleasant to some people here to consider that Apple use Blackberry's kernel OS for a major feature.

Edited 2014-03-13 08:51 UTC

Reply Score: 1

-pekr- Member since:
2006-03-28

It's not pleasant, it is a fact ...

Reply Score: 2

henderson101 Member since:
2006-05-30

Well, no. It's a streaming client with feedback. There's nothing that specifically ties that to Apple CarPlay nor is Apple CarPlay specifically tied to QNX, other than on a protocol implementation level. It's just a protocol though, and the QNX player is just a "dumb" terminal. I suspect that QNX was the initial target *because* it is almost the most common in Car OS, not for any specific technical reason.

Reply Score: 3

zima Member since:
2005-07-06

BB market share is mostly legacy devices, don't imply it's the same...

Reply Score: 2

End of an era
by ingraham on Thu 13th Mar 2014 03:13 UTC
ingraham
Member since:
2006-05-20

My company started using QNX 4.22 in 1995; we were using Photon 1.0 Beta. I didn't join until 1997. Photon was much more stable by then. I oversaw the transition from QNX4 to QNX6... and the move away from our PC-based system to a PLC-base. Killing Photon is rough; we'd have to redo quite a bit of work to move an existing system on to a new framework. Killing self-hosted developments absolutely ends QNX's singular advantage over all other commercial RTOSes. At this point, it would actually be EASIER to port our software from QNX 6.3.2 to some variant of Linux rather than 6.6.

Still, I understand. They're going where the market takes them, which is a good thing from a business perspective. I wish them well, even as my involvement with QNX dwindles away.

Reply Score: 6

RE: End of an era
by BlueofRainbow on Thu 13th Mar 2014 12:04 UTC in reply to "End of an era"
BlueofRainbow Member since:
2009-01-06

Given the breath and magnitude of the changes, it is strange that marketing did not ask for a ".0" number and won the internal debate - i.e. going to 7.0 rather than just 6.6.

I don't know the under-pinning of Photon. However, its design goes back to 1995 and it may not be aging well technically - gestures, touches and virtual keyboards as means of interaction with the user were not available even on research computer systems back then.

Dropping self-hosting may reflect the reality of software development at this time - via a virtual machine running as a guest on a Linux or Windows host.

Also dropped is the self-booting CD/DVD distribution - so it will be even harder to give it a try - like it was possible with the QNX Demo Floppy!

Reply Score: 2

RE[2]: End of an era
by tylerdurden on Thu 13th Mar 2014 17:42 UTC in reply to "RE: End of an era"
tylerdurden Member since:
2009-03-17

gestures, touches and virtual keyboards as means of interaction with the user were not available even on research computer systems back then.


Not quite. There have been commercial products using touch since the 80s, or even further. I have actually seen QNX systems running a touch interface in the 90s (industrial control application).

Reply Score: 5

RE[2]: End of an era
by ingraham on Thu 13th Mar 2014 22:50 UTC in reply to "RE: End of an era"
ingraham Member since:
2006-05-20

...it is strange that marketing did not ask for a ".0" number ... i.e. going to 7.0 rather than just 6.6.


Well, it would have to be 8, as odd numbers are reserved for internal development. Point well taken, however. I imagine the reasoning had to do with the microkernel itself; since the major core components didn't really change, they still think of it as the same "OS." You're right that from a marketing perspective jumping to 8 would have a lot of advantages, not least of which is that they've been on 6.x for well over a decade.

Photon... may not be aging well technically


I'm sure you're right. Especially with the difficulty of upkeep with the rapid pace of change in graphics.

Dropping self-hosting may reflect the reality of software development at this time - via a virtual machine running as a guest on a Linux or Windows host.


Yes, but having your dev tools on your target gave you an awesome advantage.

Also dropped is the self-booting CD/DVD distribution - so it will be even harder to give it a try - like it was possible with the QNX Demo Floppy!


Indeed.

Reply Score: 3

RE[2]: End of an era
by Megol on Sun 16th Mar 2014 15:49 UTC in reply to "RE: End of an era"
Megol Member since:
2011-04-11

<snip>
I don't know the under-pinning of Photon. However, its design goes back to 1995 and it may not be aging well technically - gestures, touches and virtual keyboards as means of interaction with the user were not available even on research computer systems back then.


It is basically managing a view space consisting of layers of sorted rectangles where each rectangle can consume some messages. The screen and input devices and be mapped into rectangles too.
The result is a very clean concept for a windowing system as messages both for rendering (travels from the redrawing rectangle(s) towards the screen rectangle) and input (travels from the front backwards) are united into one very clean design. That is the concept, an implementation is likely to do this in a more efficient manner.
Note that this can be extended into multi-screen support and per-user console support (input & output) almost for free.

It shouldn't be too hard to support a compositing design with the same basic layout but I guess they know what they are doing...


Dropping self-hosting may reflect the reality of software development at this time - via a virtual machine running as a guest on a Linux or Windows host.


Yes but would self hosting be hard to support? QNX is, after all, a full POSIX system.

Reply Score: 1

RE[2]: End of an era
by zima on Wed 19th Mar 2014 23:57 UTC in reply to "RE: End of an era"
zima Member since:
2005-07-06

I don't know the under-pinning of Photon. However, its design goes back to 1995 and it may not be aging well technically - gestures, touches and virtual keyboards as means of interaction with the user were not available even on research computer systems back then.

The history of touchscreens goes much further back than 1995:
http://en.wikipedia.org/wiki/Touchscreen#History
http://en.wikipedia.org/wiki/Multi-touch#History_of_multi-touch
http://cds.cern.ch/record/1248908?ln=en

Reply Score: 2

It might make CarPlay look and run better...
by Sabon on Thu 13th Mar 2014 19:56 UTC
Sabon
Member since:
2005-07-06

It might make CarPlay look and run better...

Reply Score: 1

Comment by brynet
by brynet on Fri 14th Mar 2014 01:21 UTC
brynet
Member since:
2010-03-02

"We no longer ship a self-hosted version of QNX SDP (that is, you can no longer use the QNX Neutrino RTOS itself to develop programs). We also no longer ship a bootable x86 CD."

So, no more installing onto a spare box and using the non-commercial license.

Reply Score: 3