The newest version of QNX Photon microGUI will enable multi-layered graphical interfaces which allows embedded developers to create sophisticated GUIs that display a mix of static and dynamic content. The multi-layer support will be available in the latest rollout of QNX Neutrino RTOS.
New QNX Multi-Layering API Enhances GUI Capabilities
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Sounds like transparency to me. What’s the big deal?
I know where I work we use a lot of qnx but does anyone else use it?
QNX couldn’t hold a candle to Windows CE and you all know it !
Oh, you mean QNX holding a candle to the new WinCE-based BMW because WinCE can’t turn on the lights and drive in the dark?
Please that whole article was basiclly MS bashing because BMW is to damn lazy to test their software.
You really have to be kidding yourself if you truely believe that QNX is dead. QNX is one of the few leader operating systems in the embedded market. A large parts of the car you cross on the road each day runs QNX, it’s used in aviation, industries and so many other places it’s _far_ from being dead.
QNX got a huge lead in terms of performances and stability over WinCE, and at least it passed the Fail-Safe certifications.
that the Linux community seemed for the most part to despise QNX. The desktop could use some work, but there are a lot of very nice things about that OS.
Oh, well. The Open Source Revolution seems to be working, so maybe they were correct in turning their noses up at it.
QNX has some really nifty features. However, it has many limitations that really keep it from being anything more than an embedded OS.
>>> Rayiner Hashem:
“[…] it has many limitations that really keep it from being anything more than an embedded OS.”
Could you please elaborate on this. I don’t see anything in QNX’s architecture that limits its use to the embedded domain. I would say that the opposite is true – you could turn QNX into a nice desktop environment with little changes.
>> it has many limitations that really keep it
>> from being anything more than an embedded OS.
> Could you please elaborate on this.
AFAIK QNX doesn’t implement stuff like virtual memory.
That simply wasn’t necessary for the markets QNX is
targeted at. (Could be that I’m out of date.)
The one thing i love about QNX is that they moved stuff
like drivers out of the kernel. And they still got better
latencies than any other OS I know about.
I simply hate it when fe. I try to use a SSH connection
to a Zaurus that already is disconnected and both Linuxes
crash hard. No driver should be able to do this, no matter
how buggy it is.
Hehe. I’m afraid veryone THINKS virtual memory is necessary, but in reality, it isn’t. On todays PC’s with > 256MB RAM, operating systems could drop virtual memory. “Normal” office work doesn’t need virtual memory anymore. Digital media (video/sound)editing cqn do without too. Moreover, this would save some memory and kernel time needed for the implementation of virtual memory.
Servers and geek users do need virtual memory, however.
You clearly have no idea what you’re talking about..
Try creating a huge 3d model in a memory-eating program like Maya, or editing a 1600×1200 image (+ z buffer image and other stuff needed for editing 3d images) in some image editor..
Man, you really do not want to try that without virtual memory and a fast HD
(altough you won’t be able to do such things with 256 MB either
Virtual memory has more advantages than just being able to swap memory pages to the HDD when space get low. Virtual memory provides the same address space for every process running on the operating system. For all intents and purposes the machine looks like it has a flat address space in the GB range without applications having to know the in and outs of the systems particular memory layout. Moreover it allows for techniques such as memory mapping where a virtual memory address doesn’t correspond to a physical memory address but is merely an abstraction into an arbitrary data source and/or destination.
>>> jeti:
“AFAIK QNX doesn’t implement stuff like virtual memory.”
Not true. While QNX 4.x already supported virtual memory addresses it did not support paging to hard disk. With QNX 6.x this limitation no longer exists. You have virtual memory space as well as hard disk ‘swaping’ (although it has to be enabled by your application).
It is paging that QNX doesn’t impliment (although the x86 kernel does have pagefile hack that the GNU compiler tools use so they can chew up the RAM they are used to chewing up). Virtual Memory is a pretty poor term all around.
1) Rich: I DO know what I’m talking about. I’m right now running a simple 12-hour simulation consuming 500MB of my 512MB RAM. Rendering is about the same thing, but 3dsmax and maya have indeed the tendence to use as much memory as they can. Well, depending on your profile, you just need more ram or a good workstation and perhaps use – indeed – virtual memory, but then you’re clearly not a normal office user (about whom I was talking above). Remind that MOST computer users are just simple office users.
2)Anonco: virtual memory (swapping parts of memory to disk) is not the same as virtual adressing, which OF COURSE is a very interesting thing. But you don’t need virtual memory to support virtual adress spaces (unless you’re a workstation user).