Today we host an interview with Chris McKillop, software engineer of QNX Software Systems. Chris is probably the most well known QNX engineer as he is always helping out the QNX community in forums, IRC etc. In this interview we talk about the embedded space, the competition but also about the QNX Momentics 6.2.x release (preview with screenshots), a free OS to download and try out on your PC (~260 MB ISO image, ~1 GB primary partition required for installation).
1. Some developers swear by the “real time” kernel architecture, while others prefer more traditional kernels. What is your belief? Can a real-time kernel do the same kinds of jobs as well as a monolithic kernel, and vice versa? Or some things are better done with a real-time kernel and other without?
Chris McKillop: Actually, a real-time OS *can* use a traditional monolithic kernel. The real question is whether a real-time microkernel – such as the one used in QNX Neutrino – can do everything that a traditional monolithic kernel is capable of.
Short answer: Yes.
Of course, microkernels and monolithic kernels both have their strong points. Monolithic kernels, for example, must either use a single “big” kernel lock (BSD) and suffer the consequences of poorer scalability or use many finer-grained locks (Linux) and suffer the consequences of complex system design, such as potential deadlocks. With a microkernel you don’t have these problems since the kernel does very little and uses existing thread-level scheduling to naturally achieve scalability and parallelism. Monolithic kernels, on the other hand, have an easier time sharing some resources – buffer caches, for example – since all of their drivers reside within the kernel. Nonetheless, these kernel-space drivers also make kernel faults more likely.
In my experience, a microkernel will do anything a monolithic kernel can, but with added benefits. For instance, it lets your system recover gracefully from faults in drivers; it also allows system services to be upgraded dynamically. What’s more, a microkernel lets you do system-level coding and debugging at the process level. Not having to reboot on each test-debug cycle makes for much shorter development times.
Regardless of whether you use a micro or mono kernel, there is a lot of value in designing your system with real-time in mind. Doing this forces you to be more aware of what is going on since you are always trying to achieve the lowest possible overhead and the fastest interrupt-servicing times. Sometimes, standard real-time schedulers – FIFO, Round-Robin, and so on – aren’t the best options for a desktop system, but you can add in other types of schedulers without compromising the system’s real-time capabilities. For example, in an OS that supports multiple scheduling types, you can keep your drivers running with a FIFO scheduler while moving your applications to an adaptive one.
2. What are the biggest obstacles today to “make it” in the OS consumer market? Hardware compatibility, hype, money and focus, app support, something else or all the above?
Chris McKillop: It’s important to remember that QNX targets the embedded market. Our customers build networking equipment, medical instruments, in-car telematics systems, and other mission-critical “stuff.” So while we do end up in some cool consumer devices, we’re not targeting the desktop.
To survive in the embedded space, your technology has to be very, very solid. For instance, people use QNX Neutrino in systems where just a few seconds of downtime can mean loss of life or huge financial penalties. Being “good enough” doesn’t play well in this space.
The desktop OS space is very different. To attract users, you need excellent – and very aggressive – marketing. To keep users, you need lots of third-party software that people can buy in their local computer stores. Achieving all of this takes very deep pockets. In fact, as long as you have strong marketing and a large pool of applications that people can buy, you can get by with a “good enough” OS.
Consequently, I don’t believe that any single vendor marketing their own OS can make a real dent in the consumer market. Instead, find a niche, work it hard, and let Apple and Microsoft duke it out for Mom ‘n’ Dad’s computer. 🙂
3. The Linux kernel has made strides lately in the embedded market, mostly with the strong presence of MontaVista (after the demise of Lineo). Do you see Embedded Linux as a serious threat to QNX and to other “traditional” RTOSs in a similar manner as Linux is threat to UNIX today – or do you think there is space for everybody? And the real question truly is, does Linux scales as well down as it does up?
Chris McKillop: The embedded market is an interesting place. The vast majority of embedded systems shipped in the world run an OS written in-house by the people who built the system. This is where open-source OSs may gain some traction. Instead of maintaining their own in-house OS, a company could turn to established bases such as NetBSD and Linux. In fact, I honestly think that NetBSD is doing far better in this space than Linux, but since there are no marketing machines behind it, you don’t hear about it in the press.
Can NetBSD or Linux be stripped down to a smallish embedded device? Yep. Does it happen magically? Nope. It requires work and maintenance on the part of the company using those kernels. Does the OS become an RTOS because it now smaller? Nope, it is fundamentally the same general-purpose architecture, but with less stuff.
Really, choosing something like Linux or NetBSD as your base means you are signing up to be responsible for maintaining your embedded OS. And when you do this, it doesn’t take many days of work to eliminate the cost benefits of going with a “free” alternative. In comparison, a commercial RTOS tailored for embedded development can let you focus on what’s important – getting product out the door.
4. What new features are you working on for the future of QNX Neutrino RTOS? What most clients are asking for these days? Have the demands on RT embedded OSs changed through the years? If yes, in what way?
Chris McKillop: There are lots of cool features coming up in the next release. The Photon microGUI, for instance, is getting an internal face lift to make it scale even smaller than it does today. Beyond that, I’m not at liberty to say – though I will point out that we’re about to release a new power management framework for mobile devices. Existing power-management models tend to impose rigid power policies on your system. In comparison, our framework offers a lot of fine-grained control over exactly when peripherals are powered up or down. This is something that our in-car telematics customers are particularly excited about.
As to your second question, demands on RTOSs are definitely changing. Today, customers want RTOSs to be more and more like general-purpose OSs but without sacrificing real-time behavior. We’ve got customers using QNX Neutrino on 4-way Xeon boxes with 6G of memory doing massive amounts of network data processing. That isn’t something your dad’s RTOS – or most RTOSs today! – could do. 🙂
5. What can you tell us about the future of QNX Momentics? There has been only a small update over 6.2.0 in 1.5 years, and we have heard nothing regarding any future plans for this desktop version of QNX. Has the ‘trick’ of marketing QNX Momentics as the development platform of QNX Neutrino RTOS paid off?
Chris McKillop: Actually, the commercial release of QNX Momentics Professional Edition 6.2.0, our Eclipse-based development environment, began shipping just over a year ago, in June 2002. Since then, we’ve added many enhancements. For instance, version v6.2.1, which shipped in February, introduced the system profiler, a very slick visualization tool that lets you pinpoint subtle errors or performance bottlenecks in your application. We’ve also added a lot of other things to make day-to-day development work easier: faster incremental compiles, better tool integration, human factors improvements, and, importantly, a large number of board support packages – these are critical to giving embedded projects a jump start.
There are really two versions of QNX Momentics: 1) the Eclipse-based Professional Edition, or PE, which provides a graphical IDE and runs on Windows, Solaris, or QNX Neutrino development hosts; and 2) the Noncommercial Edition, or NC, which provides GNU tools and runs on an QNX Neutrino host. The important thing to remember is that QNX Momentics is the development environment, and QNX Neutrino is the OS. So, when you download QNX Momentics NC 6.2.1, you are actually installing QNX Neutrino on your computer and then installing the QNX Momentics environment under Neutrino.
If you go to the Product Evaluation section of www.qnx.com, the difference between PE and NC is nicely spelled out.
As for improving the development environment, yes, there is work going forward in this area all the time. A good percentage of our customers use the self-hosted environment for their QNX development. Since it is built using the same components that our customers embed in their systems, anything we do to make their systems better also makes the self-hosted environment better. There has also been a lot of work done to enhance SWT on the QNX Photon microGUI to make the Eclipse IDE run faster and more smoothly.
6. I remember when the QNX RTP (Momentics’ old name) was marketed as a development platform for your Internet Appliance initiative. Today we know that the IA market never took off. Why was that, in your opinion?
Chris McKillop: The RTP, the precursor of today’s QNX Momentics NC, was never intended for the IA market only. Like NC, it was for anyone who wanted to either learn about QNX technology, use it at home or school, or prototype systems without having to get a purchase order signed by their boss. Mind you, QNX was doing a lot of IA marketing at the same time; in fact, we were built into a number of IA devices – 3Com’s Audrey, for instance. That would account for some of the confusion.
As to why the IA market didn’t take off, a number of reasons came into play. Price was a huge factor: PC prices were dropping like a rock and PCs offered more functionality. It’s also very hard to sell benefits like lower energy costs and taking up less space to people who drive fuel-guzzling cars and live in huge houses. 🙂
Interestingly, many of the same features that made us popular in the IA market also make us popular for automotive. In fact, one reason QNX Software Systems has done so well in the tech downturn is that we never made the mistake of focusing on just one market segment. Some of our competitors put most of their eggs in telecom, for instance, and are still trying to recover.
7. We talked about the Linux competition in the embedded field, but what do you think about the Windows competition? Microsoft offers already 2-3 micro versions of its Windows OS for devices and they try to gear their OS mostly for PDAs, writstwatches, and phones. Does QNX Software feel any danger from Microsoft’s presence in the field?
Chris McKillop: Anyone who fails to respect Microsoft’s power and position does so to their own detriment. But unlike the PC space, Microsoft isn’t selling to consumers anymore. So marketing alone isn’t enough, you need solid technology. And that’s where QNX Neutrino excels, hands down. Of course, Microsoft has good tools (leveraging their experience from Visual Studio) and lots of ties to hardware vendors from their years in the PC market. But even there, we are catching up very quickly with QNX Momentics. Because it is based on Eclipse, our customers are beginning to leverage the many third-party tools being developed for the Eclipse platform, such as Rational’s ClearCase Eclipse plug-in.
One of the nicest things about the embedded space is diversity. I seriously doubt that any one company will ever achieve the >90% market share that Microsoft currently enjoys in the PC space. And, in my opinion, that’s a good thing for everyone. Competition breeds innovation. Imagine how much better a certain large company’s web browser would be today if Netscape still had 30-40% of the browser market…
Chris is a friend of mine, so always glad to see him hitting the pavement for QNX.
>> I seriously doubt that any one company will ever achieve the >90% market share that Microsoft currently enjoys in the PC space. And, in my opinion, that’s a good thing for everyone. Competition breeds innovation.
>>
Good stuff. You can say that again 10 times over!
Smart man. I wonder how influential he is in QNXs decision making. I just wish more people like him existed…
Too bad I can’t download it using a download manager. Everytime I try it complains of some server issue.
Please, please, put a direct link on the site instead of all this fancy php redirection crap.
Apart from that, its a pretty good operating system. Now, if they can get a more desktop orientated scheduler, that would finish it off nicely 😉
Download it from a mirror then. QNXZone, Bittorrent and OpenQNX all are mirrors.
Excellent choice to interview Chris. He’s definitely one of the top OS engineers out there and a very modest person who should be recognised as one of the top engineers of the OS world of today.
Another older interview with him can be found here: http://www.qnxzone.com/readmore.php?news_id=22
CDM has been the main force behind the development of QNX on PDAs (Compad Ipaq and also the Sharp Zaurus).
Az Jim
If you go to the download section on <a href=”http://www.openqnx.com“>OpenQNX, you should be able to find more information on getting a fast download.
Good interview, liked it very much. QNX seems nice but I do not like to download Operating Systems too much anymore, does anyone know where a boxed retail version may be available?
>In my experience, a microkernel will do anything a monolithic kernel can, but with added benefits. For your
>system recover gracefully from faults in drivers; it >also
>allows system services to be upgraded dynamically. >What’s
>more, a microkernel lets you do system-level coding >and
>debugging at the process level. Not having to reboot on >each test-debug cycle makes for much shorter development
>times.
Oh..no not again flashback to the Tanenbaum (Minix) vs. LInus (LInux) kernel discussion.
http://people.fluidsignal.com/~luferbu/misc/Linus_vs_Tanenbaum.html
About QNx:
I use it 6.0/6.1 for some time now and always enjoyed using it but i have never used it for more then just playing around (Quake) and surfing with Voyager and Mozilla, wrote some words with Abiword. Made some network (nfs) connections and shared some mp3s but soundcard was never detected and i find that strange for the embedded market you have to have sound support so where are all the sounddrivers?
A REAL Desktop is also a must, icons, desktop menu etc.
Wallpaper function works already
As a system for PDAs for me its not an option, openzaurus and gpe are much more mature, have much more applications are more spread and are free/open!
So – what sound card/chipset do you have then? Gotta provide some details if you want help when something doesn’t work. And I agree that as an end-user system, OpenZ and GPE are much further ahead of eQip. But it’s ment to be a system for developers to develop on and play with currently. Which is does very well.
Not really. We once had a option to ship you a CD copy of the download edition but I don’t think we do that anymore.
It’s always great to see a QNX article or interview. Good educational stuff with a lower BS quotient than most – after decades in the market, less delusional than the newcomers.
cdm rocks too.
>So – what sound card/chipset do you have then? Gotta provide
>some details if you want help when something doesn’t work.
I appriciate that..
Its a Crystal CS4280-CM chipset and its not reconized in 6.0 or 6.1. I have searched for some answers on google and visited #qnx but nobody could help me there at that time (its about 4 months ago) the discussion ended with some guys and a former QNX developer who stated that gcc sucked but was used under corporate pressure and LInux and all open-source Os’s had unclean and buggy code..nicve guy btw.
I love IRC.
>OpenZ and GPE are much further ahead of eQip
You missed to most important point, its not free/open.
Commercial operating systems are from the past i really believe that the future Os’s are free/open. You will merly pay for using applications on the platform or the datastream. So when will you open QNX?
That he (Chris) sounds almost like JLG back in 1998-9.
Does anybody know if it’s possible to run QNX Neutrino RTOS 6.2.1 on VMware Workstation 4.0?
I installed it flawlessly creating a vm with typical settings, but it doesn’t boot, it creates error with vcpu-0… I know the OS isn’t supported by VMware, but it will be very useful if I could run it inside of Windows.
What about eQip isn’t free and open? You can get all the drivers and all the Apps in source from our project page. You can contribute back packages. The joy of using a ukernel is that nothing is in the kernel and you can extend it without having to have the source to that small part in-hand. Albeit, it is only free for personal/non-commerical use like our NC download. Which shouldn’t be a problem for you, as a user, hacking your PDA for fun.
http://qnxzone.com/eQip/
As for your audio, it doesn’t appear to have a driver.
You could always try writing one. MikeG and Igor have both done ports from Alsa to QNX’s ALSA-like audio system with success. There where a couple of others as well, but thiers got to completion.
What I’d really like to see from QNX is a VMWare graphics driver (VMWare only provide drivers for Windows and Linux) — this would let people run QNX at a decent speed from within Windows, rather than having to sacrifice a primary partition, which is quite a big deal for most people. At present, QNX NC is very sluggish under VMWare using the default graphics output (as Windows and Linux are until you install the VMWare Tools drivers). Seems like a relatively easy way to get more people using QNX stuff…
Better ask Mr. McKillop why his company never fullfilled its promise to make the QNX source “accessible” as they promised back in year 2000. This outdated cvs.qnx.com surely doesn’t count. Let me give you a quote from their press release (http://www.qnx.com/news/qnxis.html):
“Besides making QNX free for non-commercial use, QNX also provides source code for most components – drivers, protocol managers, and so on. Only the kernel and other core portions of the QNX platform remain protected, giving OEMs the benefit of building their systems on secure technology with clear IP ownership.”.
General acceptance of QNX will be predicated on the royalty fee structure.Linux’s encroachment in embedded system products is partly due to its lack of royalties as well as low initial cost. You can test drive QNX all you want, but watch out for the distribution license cost.
WRT licensing and Linux, this is not a valid comparison. I’d be willing to bet that QSS considers WindRiver’s VxWorks as its primary competition, and not Linux. Yes, Linux is a blip in the embedded/RT world – however it has only penetrated “cheap” devices, and nothing of a critical nature.
If you then consider the licensing scheme of a WindRiver, or even Green Hills it is evident that the QSS is will within what the market bears.
That said, you must look at the hardware environment where QSS is operating; deeply embedded. We’re not just talking about some OTS x86, ARC, or PPC platform.
So what exactly is: “General acceptance of QNX”? There is no such thing as general acceptance in the RTOS/embedded world. Everything is about functionality per project, or application and nothing about the number of “general” users.
…enough ranting.
Kit
>”cheap” devices, and nothing of a critical nature.
Sure..keep on dreaming..embedded Linux will even even power the next mars rover….see http://www.timesys.com for more info, but then again that is not real mission critical is it?
http://slashdot.org/articles/00/12/13/2012224.shtml
http://www.metrowerks.com/MW/Develop/Embedded/Linux/default.htm
http://www.elinos.com/
http://www.lynuxworks.com/products/bluecat/bluecat.php3?ck=54
http://www.lynuxworks.com/products/lynxos/lynxos.php3
http://www.timesys.com/
List goes on and on…
“If one process is using a kernel facility, no other process can execute until the process finishes or waits on something.”
Can you imagine what will happen if 2.6 will come up?
Lest face the future is bright the future is Linux!
>personal/non-commerical use like our NC download. Which >shouldn’t be a problem for you, as a user, hacking your PDA >for fun.
>http://qnxzone.com/eQip/
Thanks, i did no know about the free part.
I read that Zaurus support is underway, pitty i am
very eager to try it..
>As for your audio, it doesn’t appear to have a driver.
>You could always try writing one. MikeG and Igor have both >done ports from Alsa to QNX’s ALSA-like audio system with >success. There where a couple of others as well, but thiers >got to completion.
I have no experience with writing audio drivers but i am in the position the buy the supported hardware so that should no longer be the problem.
I hope you guys will follow the ecos way..i like it that way.
Anyway. Thank you for your time you have put in the article and the time you have put in QNX and its maturing.
Yep, TimeSys will be on the next Mars rover. However, take a look at what that actually means. A very heavily modified version of Linux that uses non-GPL loadable modules to impliment the systems required to make Linux even begin to look like an RTOS. It is not stock Linux by any means. I think that is important to point out.
In your install of QNX go look in /usr/src. If there is a specific component you need source for feel free to ask your support rep. about getting access. Is there some part in particular you are interested in?
>Yep, TimeSys will be on the next Mars rover. However, take a
> look at what that actually means. A very heavily modified
>version of Linux that uses non-GPL loadable modules to
>impliment the systems required to make Linux even begin to
>look like an RTOS. It is not stock Linux by any means. I
>think that is important to point out.
Absolutly! Linux is not a real RTOS but it is beginning to
mature..
QNX and vxWorks are leaps ahead but Emmbede Linux is catching up very quickly.
funny thing..
http://www.windriver.com/news/press/20030606.html
Could it be the vxWorks Java layer works togheter with
the LInux/Java layer?:) I remember having joking about the Linux-mars version that with Alan Cox when i signed him up for the cdrom that was aboard on the mars-rover..i forgot the URL however
( things can go quickly these days..
:s/Emmbede/Embedded
Sorry but, as everybody noticed already, my native language is dutch..
While it’s been ages since I’ve played with QNX (3yrs+), I do remember that I has ports of apps like AbiWord and several other seemingly less important ports of games like Quake. I wouldn’t be as curious as this if some Joe FooBar did this on the weekend but the fact that these are available in the package manager (or alike) shipped with QNX personal surprised me.
Is the port of quake designed to show the capabilities of this realtime os, eg. as a benchmarking tool or do some of us folk reboot into QNX for a multiplayer deathmatch?
Regardless of this, Chris seems like a very knowledgable engineer. Excellent choice of person for an interview, Eugenia. Also, congrats to Chris and the QNX team. It looks like this platform has potential to gain a bigger slice marketshare thanks to its technical brilliance rather than marketing. I’ll have to have a look at it. I only wish I had a need for a realtime os… hmm…
I’m running numbers for 5 days training, 5 developer seats for 5 years, 500 unit binary distribution plus full source code. It doesn’t seem to matter who provides the OS — while the types and degrees of financial burden vary, they all want to bite you for as long as possible.
Some vendors want big money up front only once; others want annoying fees yearly forever; Wind River (VxWorks) wants both.
QNX looked very good until we got to the $$$/unit distribution royalty. Full source availablity varies. If you are Cisco and run a rebranded QNX in all your products, you may get the source, in part because you propagate fixes back to QNX. Otherwise, Neutrino source costs a disturbing amount of money.
It’s a juggling act, I suppose.
—
I see your name one the front page of osnews. Great article. Sounds like you’re still having fun at QNX.
I don’t know where you’re located, but I can’t imagine that the licensing costs you would incurr outstrip what you would pay 5 developers for 5 years if they had to recreate the features or twist linux to what QNX does for you out of the box.
If you’re in the US 1 developer’s salary would be around 80 to 90k, total cost to the business (w/ health insurance, taxes, etc) would be around $120k to $130k per year. So for just one year it would be $600k to $650k.
I bet your descision to go with QNX will save you more than a year in your development efforts….I don’t know what the full source the QNX costs, but is it much more than $650k?
Has anyone heard of anyone using QNX as the OS installed on a cluster of special purpose PC nodes, like the kind used to drive tiled display walls? Usually Linux clusters are used for this, but they suffer frame rate challenges. Lots of that is due hardware data transfer limitations, but it seems that more than a little may be affected by software latencies.
In particular, does anyone know if Chromium (lets Linux, Irix, and Windows openGL based programs to run on tiled display walls — http://sourceforge.net/projects/chromium/ ) can be made to run on QNX? I’ll bet QNX doesn’t have drivers for cards like Quadros or high end ATIs (but I’d love to be shown to be wrong). But aside from that, what else would stop one from using QNX for visualization clusters, which strikes me as a hybrid between a workstation and an embedded application. (they don’t do much other than process streams, and you don’t want the nodes to ever slow down for any reason or crash).
— Ed Boyce
I am few days away from trying eQip on my iPaQ, even if it’s not yet fully usable. But I am wondering if the eQip project is just a “toy” developped by some developpers at QNX or could it become a serious project for QNX ? It fits well the embedded market that Chris is talking about but it was never clearly stated what is the future of that project.
This is always an interesting point, the linux community has been “conditioned” to a great degree to always demand source. They do this for a very good reason, there is no other way to debug anything in Linux without source, so it becomes natural to expect it and demand it when it is not available.
QNX is a very different beast when it comes to actually needing source to do things. The entire operating system leverages the IPC mechanism to it’s fullest advantage. That means that with the docs that are available when you install even the NC version of QNX, you have all the tools and docs that you need to extend the OS and in many cases debug parts. Just about everything in QNX is a resource manager, and it’s dead easy to write your own to extend the operating system. To extend Linux in a similiar manner would require source, and recompiling the kernel, and most likely a reboot. Debugging anything like that in linux is very difficult, you just can’t set a breakpoint in the kernel and not hose the entire system. In QNX, it’s a resource manager, it runs in user space, you can use standard debuggers and techniques (even printf, which would be impossible for a debugging a driver in linux).
The bottom line is that in linux you need source to do anything. In QNX you rarely if ever need source.
The source on cvs.qnx.com may not be 100% recent, but there is more than enough there to give you all the “undocumented” information required to do just about anything in QNX.
If you still think you need source, come talk to me or cdm on IRC (#qnx channel on irc.joher.com) and I am confident that once you explain what you want to do, we can show you how to accomplish it in QNX without the need for any source from QSS at all.
There has been much talk about licensing for QNX and the costs related to this especially when compared to Linux. So I’d like to give you my perspective as someone that has just recently done an embedded project with QNX.
The project was quite simple, my employer needed to create a device capable of collecting data and delivering it reliably to their servers via the internet. The connections would be satellite, or wireless, so the device was at the basic level a communications controller. It would talk to “dumber” devices to get the data, and then manage all the intricacies of delivering data reliably over the internet and via wireless from remote locations. Since the majority of the installations would be 2-way satellite internet, it also had to act as a firewall, providing similiar service as you’d expect from a personal firewall for your DSL or cablemodem service.
We started off with an Advantech FWA-240 ( “http://www.advantech.com/products/Model_Detail.asp?model_id=1-DEN4J… ), which had the firewall component already there as well as Linux.
We tried to modify the linx setup, and it turned into a nightmarish experience. RPM packages had been installed and then part of them manually deleted, it wasn’t running the latest release of the packages or kernel either. So just adding in the missing parts of the package management tools became a difficult task, and we hadn’t even started to make the mods for embedded or our application. Linux was not terribly “flash friendly” and was writing to the CF module all the time, which was bad, and modifying much of that required a new kernel compile, and possibly modifying/customizing some standard Linux apps/utils.
We gave up. We too QNX, and I had the thing booting in 3 days (boot strapping QNX into this was the challenge because there was no vid hardware). Three weeks later we had an alpha release of our application running. If we had stuck with Linux we might have gotten it to stop writing to the CF by that time.
QNX easily saved us 6 months of development time, not to mention that it saved us the costs of maintaining all those custom versions of Linux and the utils that we would have had to make.
Runtimes for the units were <$120US per unit in VERY small quantities. We could easily afford this in our unit costs when we sold the devices, and the time and money saved by choosing QNX allowed us to land a major contract that used the devices.
With Linux we’d have spent more, not had as stable a product, and might not have gotten the contract.
What do you mean a real product? Do you mean is someone making hardware (A PDA) based on it? Do you mean is it being activly maintained and moving forward?
Does QNX support the SoundBlaster Audigy at all? 🙂
Adam
The following link will show you exactly what SoundBlaster cards are supported. It doesn’t look like the Audigy is supported, but feel free to port over the driver
.
http://www.qnx.com/support/sd_hardware/platform/audio.html
>>>But I am wondering if the eQip project is just a “toy” developped by some developpers at QNX or could it become a serious project for QNX ?
QNX uses the qnx/ipaq quite often to demo to their telematics customers. Both use quarter vga lcd screens (telematics just uses the screen sideways in landscape mode). Both use some sort of strongarm chip.
http://www.summermeat.net/topics/samples/auto4.jpg
http://www.summermeat.net/topics/samples/ipaq_demos/
It’s not a toy for QNX — it’s serious business for them.
The “toy” aspect is your everyday pda apps (but you can always try to port them from various linux projects). But QNX do have customers inquiring about the usage of qnx/ipaq in industrial settings. Put the ipaq into an OEM rugged case, and you can use it on the factory floor. You will never see these devices in the general public, but they will be around.
http://h18000.www1.hp.com/products/quickspecs/10971_div/10971_div.H…
.. is a L4 kernel with a Linux frontside (and it´s complely Opensource)
I know this isn’t a QNX support forum, but I thought I’d mention this anyway… If I can’t figure out this problem myself, I’ll look for support in all the right places.
I went ahead and installed QNX onto a 28 gig partition of a 128 gig hard drive. This is the third partition on the disk. The install went fine. However, when I reboot, grub refuses to chainload this partition. I specify the root as (hd3,2), and when I do “chainloader +1” from the grub command line, it complains about an unknown or invalid executable.
Thankfully, I can still boot off the CD and directly into QNX on the hard drive I just installed it to, but it would sure be nice to get this working properly 🙂
Note: Both Windows XP and FreeBSD (the first and second partitions on the drive) chainload just fine.
Adam
Is that partition a primary one? QNX requires a primary partition. Plus, 28 GB is a bit of overkill for QNX. You will never use more than 3-4 GB anyway…
OK, I guess, I have to explain more. I meant is it going to be developped up to the point where eQip could be use by average PDA user as a replacement of PC2002 ?
Chris already stated that they would not compete MS on the desktop market, I was wondering if eQip could take some share on PDA market. And not only for technical demonstration but also for more traditionnal use.
I cannot program myself the applications I need, so I need to rely on what QNX or 3rd party would develop… so I am hoping they will continue eQip
I would love to see it to that point someday. But today it is not ready for general PDA use.
I am now working with the QNX system for almost 5 years and I think it is one of the most stable OSs that I know of.
With the step to QNX6 some things changed here. Like Phindows is no longer running stable, some widgets in the PhAB are no longer working correctly. O.K. this can happen, if this is fixed with an update, no problem! But if you have to pay for the update, no thanks! I think this is a bad policy of QNX. Or maybe that is hust QNX Europe, I do not know that.
But I have to agree with the other people: interresting interview Chris!
i also wonder what are you goals for eqip, i would like to use it on an ipaq 3970. when will eqip be ready for it?
p.s.: why is there no QNX in the name?
eQip == Embedded QNX for Intelligent Platforms
There are other wordings people have used for the eQip monkier, but that is the one I normally quote.
It is actually an “opensource” project being done by people in and out of QSS Ltd. That is why it isn’t called QNX PDA 2003.
I went ahead and installed QNX onto a 28 gig partition of a 128 gig hard drive. This is the third partition on the disk. The install went fine. However, when I reboot, grub refuses to chainload this partition. I specify the root as (hd3,2), and when I do “chainloader +1” from the grub command line, it complains about an unknown or invalid executable.
Try “chainloader +4”
There is a project (called RadiOS), that will have a free Neutrino-like microkernel (and process manager). Written in 100% x86 assembly (NASM); currently there are about 30 syscalls implemented (out of 64).