Slashdot Mirror
Explaining The Symbiosis Between QNX RtP & Linux
Explaining the Symbiosis between QNX RtP and Linux
The Linux community and its open source approach has huge potential and helps us greatly to realize our goal
to make the QNX Realtime Platform a superior general purpose desktop and developer operating system. QNX RtP
is fully compliant with the latest POSIX standards. This, together with a X11 compatibility layer for the
Photon MicroGUI, means most Linux/Unix applications only need simple recompiles and modifications to be
ported to this new platform. So whatever Linux/Unix gains the OS will gain as well! Currently QNXStart.com already has a nice software library and Tucows is busy
building one. The freely available award winning Voyager browser is excellent and plug-ins like Flash 4 and
RealPlayer 7 are already available. Also Amiga`s Digital Environment is being developed to support the QNX
RtP and will not only provide us with a common binary identical application layer for Linux but most other
operating systems as well. The Tao Group (Amiga`s OS partner)
already provided QSSL with a wonderfully
small and efficient Java VM implementation for its predecessor QNX4. In addition QSSL and IBM are working together on a real-time technology for Java
applications under the QNX Realtime Platform.
Giving source code away freely for everyone to use does not make sense for everybody. Especially commercial
companies would not be motivated to spend years on OS development and make the fruits of their work freely
available for rival companies. With QSSL being the owner of this OS and
offering all the relevant source code to developers, developers can invest their resources while still being
protected from competitors. So QSSL offers prime advantages of both the open-source and commercial worlds!
!
One month ago the pre-release developer version of the QNX Realtime Platform was freely released for public
download freely for non-commercial use. Since its launch more than 400,000 downloads at http://get.qnx.com followed, while bringing even Tucow's servers to their knees. Additionally, QSSL has set up a developer's support network with weekly articles by QNX experts giving their
insights on programming under the QNX Realtime Platform. There are very ambitious projects for this new OS
and many exciting new developments will be announced when the time is right.
But what makes QNX RtP so special, you may ask?
QNX RtP is an excellent realtime operating system, which means that all programs are smoothly given CPU
time (according to the priority they are given in the Scheduler). This gives developers or users the option
to set high priorities for certain tasks, so that for instance a multimedia player will be quaranteed to
respond and function optimally in any given situation. The stability and realtime abilities of QSSL`s OSs
resulted in them being used heavily in fault intolerant and response critical systems, for instance nuclear
reactors, medical equipment, space craft, traffic control systems , etc.
Within the embedded market, a small memory footprint and optimal efficiency are very important issues. QNX
RtP is based on the modern and optimised QNX Neutrino microkernel. OS modules such as file systems, TCP/IP
and even drivers run as normal memory protected user processes, allowing them to be plugged in and removed
at any time without a reboot. This allows QNX RtP to be scaled down or up very easily for use in compact
Internet Appliances as well as in full blown multi-processor servers (3Com`s Audrey IA has just been
released and uses the QNX RtP at its core.). It also eliminates the need for special kernel APIs and
debuggers, greatly simplifying driver development and debugging.
Although I could continue for hours telling you endlessly about other benefits I will end this by
highlighting one other great feature which impressed most developers involved instantly, when we started our
relationship with QSSL a couple of years ago. It truly offers superior flexible and transparent networking
abilities. As an example of its flexibility; you could have a game running on one computer in a network,
while it is being controlled by a joystick on another computer within the network, and its graphical output
being displayed on a monitor of another machine again! One demonstration was of Doom running on two
connected machines to begin with; it was running on one machine, then the window it was running in was
dragged onto the display of the second machine, then it was partially dragged back, so the game was running
synchronic and seamlessly with half a window on each screen!
Something wonderful is coming. Do you want to join the exciting battle to change computing and topple the
choking monopolies within the industry? :)
Sincerely, Mike Bouma.
AmigaRing http://www.stormloader.com/amiga
Phoenix Developer Consortium http://www.phinixi.com
Just listen to this tripe about how QNX is the "best of both worlds". It amounts to this:
QNX: Gee guys thanks for all the free code.
Linux Community: No problem can we have...
QNX: NO!
LC: Can we see the source for...
QNX: NO!
LC: Okay here's some more free code though...
Honestly, QNX is just profiting from writing a kernel and then porting a bunch of free software to that kernel, while keeping the kernel proprietary. Boy that sure sounds like the best of both worlds for *them*. But for the lil' dudes we must toil and suffer. Honestly QNX is a great RtOS (in the true sense of the word) and I would love to see a RtLinux that had the scalability and power of QNX...but with the QNX attitude that will never happen...
So HEMOS how do you have vested in QNX????
You lamer...
A comment from "(Mike)(duh)" further on down claims the difference between BeOS and Neutrino is that the latter is attracting developers and BeOS isn't. This is highly misleading. But, I'm not sure Ars-Fartsica is 100% on target, either.
BeOS did have a lot of announced commercial support this time last year--from companies that to date still have shown no interest in Linux. While there are some quantifiable advantages BeOS has in media performance over Linux (and, I would submit, over QNX as well), the difference wasn't technology--nor was it that the companies didn't "trust" open source, as some BeOS cheerleaders claim. It's that Be had strong developer support. These companies wanted somebody there able to offer immediate, detailed support and corporate evangelism. Linux can come close to this sometimes, but it definitionally can't match it, simply because there is no "Linux Company" that gives you access to nearly all the great developers. "It's open, you can make the changes yourself" isn't as good an answer as "we've tracked down that bug and fixed it for you," if (and it's a big if) there's a group able and willing to do the latter. This marketplace reality is why companies target specific Linux distributions--it's because Red Hat, SuSE, and Caldera are trying to do for Linux what Be was doing for BeOS. A company may feel comfortable porting to "Red Hat Linux" because they're not doing the port alone, they're doing it with Red Hat's technical--and possibly even financial--support.
So what happened to BeOS? Be stopped providing that support for the desktop, shifting nearly all their DTS resources to BeIA.
I've looked at QNX RtP, and sure, it's pretty neat. But it is in the same position on the deskop that BeOS is--the desktop incarnation of a closed operating system whose company intends to focus their resources on the "appliance" market.
Look at the business and marketing FAQs on get.qnx.com. Do you see anything about their plans for the alternative desktop market? "It's uniquely positioned to become the premier platform for embedded devices." When they're comparing it to Linux, they're talking about its advantages to embedded systems developers:
QNX RtP isn't intended to be a desktop OS contender: it's been released as a "self-hosted development environment" for RtP-based appliances, as well as a demonstration to potential partners of what the system can do.
Just like BeOS is for BeIA now. Rah, rah, focus shift.
QNX is a bunch of great guys, from what I can tell, but don't kid yourself, folks. This is not the next great alternative OS; no operating system is going to be without serious pushing from a corporate developer.
Opening the source helps with developer mindshare, sure, but I'm not sure it'd help much; there'll be enough hackers porting Unix stuff over without it, and it's a rare profit-driven company that gives a fig about such issues--they just want to sell their products and/or services, and that requires a userbase. (If BeOS had a few million users, Corel would have been there porting already. Whether we'd want them there is another question, of course, but I digress.)
In any case, the source isn't going to be completely open in either case; both companies see their kernels as their prized product. In practice, they could use a license like the Aladdin Ghostscript one (essentially, free for most use, but commercial redistribution requires a license), but let's not hold our breath.
They seem to be saying how wonderful that all this free source code is wonderful because it takes just a recompile to work on QNX, but Giving source code away freely for everyone to use does not make sense for everybody so they don't have to give anything back.
Is this what it says? Am I confoozed?
--
Infuriate left and right
Whoever said that Linux can cure any of those things (insert "world hunger", "war", "disease", whatever else here) ? Are you smoking crack? Give me one example, one quote.
Some of us are simply not interested in closed-source software. Does that make us wrong? evil? Are we not allowed to express that opinion?
I hate to break it to you, but Linux has been ready for my desktop for years, and I don't really care about Joe SixPack. If someone else does, that's fine, but it's not like Linux is going to lose money and go out of business and strand all it's users like the Amiga did... or BeOS might.
"Free your mind and your ass will follow"
What market is that? Sure, you Linux zealots (Read: Slashdot Audience) won't use BeOS just because you cant recompile the kernel because you, in your infinite wisdom, think that it would be better with|without a specific option.
:), but the thing that bugs me most about Linux are the claims by its users that it can cure (insert "world hunger", "war", "disease", whatever else here.).
:)
Does this mean there is no market for BeOS? Hardly. I hate to break it to you, Linux is far from ready for the desktop. BeOS isn't a server OS (Hell, if you want a server OS, use a BSD), and it isn't meant to be. The market for BeOS is currently those who want an alternative to Windows on their home/desktop system, but can see that Linux isn't quite there yet. Small market, yes. Good market? well, thats something I can't comment on.
Be has a solid OS, and just because its not Open Source doesn't mean it sucks and that Be should drop into the infernal abyss.
Linux is a decent OS, it has a few things to learn (Which is why I use BSD
I grow so tired of people like that. I'v probably started a flame war here, but I'll be gone for a week so whats it matter
from http://amiga.eden.it/news/QNX-Linux -en g.html:
QNX is often compared to UNIX, LINUX and BSD. We do share the same POSIX APIs, and most code written for these systems ports easily to QNX, but the resemblance ends there. Based on 20 years of OS experience, QNX has a radically more advanced architecture. It's a massively scalable, multi-threaded, fault-tolerant, realtime OS designed for devices and computers of any type or size. QNX provides a unique network architecture where large full-service protocol stacks aren't required on each computer and devices plugged into the network are simply "discovered" automatically by other devices - all services and peripherals of the new device can then be used by any other device in the network. QNX is also the only self-hosted RTOS where the development environment and the runtime target environment are the same. And though we're not open source, we adopt an "open source" policy for hardware-specific drivers. This allows us to continually support the latest hardware advances, while still maintaining control of core technology. More importantly, it ensures QNX has a focused vision for the future.
-- "Complacency is a far more dangerous attitude than outrage." -Naomi Littlebear
They hit the nail on the head. QNX is interesting and good and cool because it will expand, yet again, the market for linux binaries!
:)
Remember, the main reason SO many people code for windows is the simple fact that in doing so, they reach some 75% of the market (or whatever it is today).
Between Gnome becoming default on Sun stations, and QNX, and all the other places (dont forget BSD!), linux binaries are reaching a seriously larger audience.
Whether individually those things are in and of themselves cool (be nice if kde AND gnome were BOTH defaults.. ie, a choice), as a whole, they are increasing the market for programs for linux.
That is how to win the war to grab programmers.
(I know what you are thinking -- who wants windows programmers. We dont. We want programmers that HAPPEN to program for windows because of its HUGE installed base).
Eventually, linux will be the OBVIOUS choice.
GPL'd web-based tradewars themed space game
I apologize if I sound rude, but his is nothing more than an advertisement. Only two or three lines has anything to do with "Open Source"(yeah, right, who here is gonna call THAT "Open Source"???), and the rest of it was them telling us how great QNX is, how well it scales, how easy it is to program for, etc., etc..
Now, if you want to review something, that's great by me - I like reviews. I don't even care if it's a review of an Evil Microsoft product, and I don't mind if you sing its praises.
But is this the forum for blatant advertising? Sure, banner ads are fine, whatever. Even "sponsorship messages" are cool - but LABEL THEM AS SUCH.
Dave
'Round the firewall,
Out the modem,
Through the router,
Down the wire,
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
I have a ten-years-old distributed system (small, just 3 nodes, but distributed) that uses an early version of QNX to remotely boot diskless front-end systems and operate them from a central computer (a 386-33MHz speed demon). The system IS nice, you can run programs on a remote system from the command line with trivial ease, and the networking (at least the Arcnet system we used for the network-boot architecture) was quite robust. (Of course, the lousy application program that we ran on top of QNX is cursed to the deepest pits, and the system has always been marginal as a result...) QNX itself is s fine OS.
o ne-calls, changing the NIC and video card 'cause LynxOS didn't like'em, and general difficulty using it, I went out and bought RedHat 5.0, and I'm never, ever going closed-source again unless forced. (BTW, the funny thing 'bout LynxOS was, all the primary development tools were GNU; gcc, gdb... Only the profiler and MetroX X server were proprietary. Now, tell me again why I paid $30K for this package?)
On the minus side, QNX (at least then) did NOT let you create a bootable floppy, something that annoys me no end. We had sufficient licences for all nodes (at $hundreds per node), but ya still needed those double-damned fingerprinted floppies to make it work.
More recently, I had a brief fling with LynxOS 2 years ago; after several days of can't-get-it-to-install-even-with-tech-support-ph
Yes, I MAY be able to get superior real-time performance out of a closed-source OS (unproven IMHO); but my systems, like MOST of those out there, do NOT stretch the boundaries of achievable performance. I'm running P5-200MHz front-end systems using VMEbus; if I need more performance, (which I haven't yet and probably won't), I'll just slap in a faster CPU card. So far, I've had to hack the NIC code once to eliminate a funky media-autosense problem, and the VME driver to accommodate the VMIC VME implementation (the latter has since fixed by VMIC; great people, I HIGHLY recommend them, AND they support Linux on their iron!); I could NOT have done this with LynxOS or QNX, and MY experience says there's every chance I would have needed to do so.
It's not ideology that moves me to say this, and not cheapness (the OS cost isn't an issue in my systems); it's Linux' better flexibility, equal or better reliability and hardware compatibility, and that source accessibility that make it my primary choice for an OS.
As a practical matter of getting my job done, Open Source wins for me. Period.
"My strength is as the strength of ten men, for I am wired to the eyeballs on espresso."
Talking about Quake and Doom to explain what QNX can do gets the idea across but compared to what QNX is trully used for, these description are pretty weak. I have seen QNX used in satellite control, telemetry, GPIB and all the wonder stuff. QNX will always have its niche guaranteed and has been around for ages, it doesn't need to make its proof like BeOS does (an other comparison which makes me uneasy).
All I can think of is they though "Hey those Slashdot kiddos only play quake, let's talk their language and avoid the serious stuff... Oh and the GPL, tell them it's not seen as good in the suit and tie environement."
A bit disapointing from them. They have a heck of a product. Also what was the heading again? Perfect symbiosis with Linux? The letter doesn't seem to coroborate this statement.
PS: Mr QNX, could you make you RTS Free CD boot from SCSI CDROMs? Somehow it doesn't look very "serious" either.
Hold on a second there, before you flame me. I love linux and I run slackware. I also love FreeBSD. And I run OpenBSD on my laptop. But there is one thing I really don't like about Linux--the design of the kernel. I dont like the idea of recompiling the kernel for new drivers. QNX is truely modular. Companies can release binary-only drivers for QNX and they work. This does not work well with linux, because the kernel changes every second.
You see, Linux started as a project to run on one computer. QNX RtP is designed by engineers who had a clear view of what they want it to be when it is done. It is well engineered. With linux, you have to change the kernel to add drivers! That is NOT the way it shold be done! This is not a problem with QNX.
All this being said I still love OpenSource, and I like linux but hate the way drivers are implemented in it. I would trade a poorly designed opensource kernel for a smaller, more modular commercial one, as long as the rest of the OS is open source. With QNX, you dont need to hack the kernel itself, because it is very small.
Modularity.
Modularity.
A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
Of course Be has a quality product - but part of the marketing of that product is a decision whether or not to open the source. Like it or lump it, The BeOS market and the linux market are basically the same - people who are willing to try an alternative OS - and most of the users in this market are using the openness of the source in their product decision.
Be could have made that decision early on to capture mindshare - they didn't - and now they are toast. The bottom line is that you give users what they want. Even if the CEO doesn't think open source makes sense, you listen to users. If users want a free car wash and a commemorative plate with each OS install, thats what you give them.
Linux is a decent OS, it has a few things to learn (Which is why I use BSD :), but the thing that bugs me most about Linux are the claims by
its users that it can cure (insert "world hunger", "war", "disease", whatever else here.).
Speaking as someone who has used FreeBSD for five years, I can tell you that most of the cool userland stuff to land on BSD in the past three years has come from the linux userbase. Gnome and KDE would not have ever come about if not for the linux craze.
My advice: learn the lesson that Be didn't - open source your OS if you want to survive. Regardless of whether it makes "strategic" sense for the company, the market for this type of product won't dedicate any mindshare to a closed source product. Without developer mindshare, you're toast.
Read the FAQ. Their opinions are stated in plain view:
Q: Why doesn't QNX provide source to the kernel and other core OS modules?
A: Because QNX developers don't need kernel source to extend the OS.
Anyone who's ever done serious work in embedded systems know the kernel source is absolutely essential for debugging, not only the application but also the kernel. All OSes contain kernel bugs. They're a pain to find and fix without source, and those of us who've been there are not going back lightheartedly. You all know this, that's why we're embracing open source. How come so many of you are now eager to jump back into the dark hole that is proprietary software?
For embedded work, there's ECOS already. It's Free Software and runs on a dozen different CPUs, with new ports coming all the time. If you want the 3D acceleration, anti-aliases graphics and macromedia player, you're probably not looking for embedded stuff in the first place.
Sure, QNX is fun. Play away. But it isn't the future.
--
--
You are a fucking moron.
Perhaps a new moderation category would be helpful:
(Score: -1, Marketing Fluff)
:)
-----
D. Fischer
ShoutingMan.com