Guest post by As of April 2010, a silent change was made on Foundry27, users with a myQNX account could no longer checkout/update their copies of the QNX SVN repositories and a vague Wiki page was created “detailing” some licencing clarifications.It appears this may have happened due to the Research In Motion acquisition, but perhaps there were other reasons. It’s amusing how at first they wanted to make the development of QNX more “transparent” and now it’s back to being proprietary, and hobbyists are no longer permitted access to the following components, as mentioned on their site:
QNX6 microkernel, userland, libraries, GUI (Photon) and other aspects of the QNX Neutrino operating system.
The only portions left available are a few drivers, the ported NetBSD networking stack, and some other community initiated Foundry27 projects. A lot of technical sites covered the opening of QNX6 development and source, but, none have mentioned this recent relapse of the project, no longer is it open to study by hobbyists.. at least in source form. Very unfortunate.
You can’t expect Research In Motion to do anything that’s smart.
That’s true.
There will be more info given within the next week or 2 on the state of QNX. But RIM is keeping things quite until it’s sure what it wants to do with QNX.
RIM has nothing to gain by meddling with this proprietary crap (even if the proprietary crap is something they own). Going your own, proprietary way pays off if you have an established position already (Apple).
If I was RIM, I would dump the other stuff they’ve got, adopt MeeGo/Android and add their proprietary email stuff on top of that. They could cut most of their R&D expenses that way.
Actually, HP could have done the same thing instead of dumping the money on Palm, but money moves in mysterious ways…
I have to disagree. RIM could potentially have a TON to gain from QNX. Using QNX as a base for their blackberry OS would put them in a great position. If for nothing else than QNX is being used as the multimedia system in a lot of cars and it would make it dead simple to integrate your blackberry device to your car. phone/car integration is key these days (or is starting ot be).
That all aside, QNX is more stable and can have a smaller footprint than the current blackberry OS.
as for HP and Palm. HP gained not only WebOS, but all of Palm’s patents, and thats HUGE!
Edited 2010-07-15 19:07 UTC
MeeGo has an in-vehicle infotainment profile as well. I have to admit I don’t know much about that, but I believe integration with those systems doesn’t require kernel level compatibility. If QNX is using X (or VNC?) in these systems, I believe MeeGo (or any other Linux) would be a drop-in replacement.
QNX used it’s own embedded graphic system Photon microGUI. No X by default if memory serves correctly.
yes, QNX uses Photon and not your standard X (and personally I love Photon).
Funny, because I think what HP did is probably the most interesting – and certainly a smart – option : get their own complete stack (metal to glass) which will allow them to make devices with unique features and an interesting interface that can differentiate itself.
Aside from not paying a ‘Microsoft tax’ and ‘saving’ on R&D costs, the Android vendors are basically stuck on the standard Windows route – very little to differentiate themselves on the software side and razor-thin margins on their hardware sales, not to mention Google dabbling in their own hardware versions to further screw them over.
Investing in Pre and R&D gives HP a shot at being a real competitor, and as Apple has shown you can do both hardware and software and make very good money with much better margins. Whether HP can capitalize on that remains to be seen – do they have the technical vision, can they get software shops interested with enough seats to make it worth porting/writing for the platform, and can they fight both Apple and Google platforms? Who knows, but they have deeper pockets than Palm did and possibly the will to try. Kudos, I can’t wait to see what they can come up with and hope it isn’t a total disaster.
And, uh, go QNX6! (To get back on topic…
)
Apple leveraged their OS X developers. Android leveraged all kinds of people who want open development. Microsoft has all the Windows fans. WebOS brings very little developer support to the table.
And Palm could have taken BeOS and made perhaps the greatest mobile OS ever. Instead, they practically gave BeOS away to a company that ended up never even using it. And sure, WebOS is nice, but HP will probably kill it in its attempt to make a viable tablet. The vicious cycle continues.
Man I wish you were CEO of some company, what a great plan. Lets dump proofed to work, has long history of devices, is well known, well liked system, that is selling rather well and move something that isn’t well know, doesn’t have any history or proofs that it’s working or selling well. Mission complete, what next? Take out all company money, go casino and bet all on red? Damn it was black
QNX is not “proven” for whatever use case RIM might have for it.
Linux, OTOH, is.
“QNX is not “proven” for whatever use case RIM might have for it”
Then why RIM has bought QNX in the first place ?
Kochise
From
http://www.engadget.com/2010/04/09/rim-buys-qnx-talks-in-car-infota…
So, RIM had an actual reason to buy QNX – it seems some nice embedded products with QNX will come out, and RIM gets the inside position to integrate with those.
I don’t see QNX going on their phones though.
Looks like you never used QNX. It’s ok, but don’t say that it’s a ‘proprietary crap’. It may be proprietary, but is definitely *not* a crap. OS itself is very old, stable, fast and standarized. It makes lot better RTOS than Linux RTOS appliances.
I tend to refer to proprietary infrastructure (when good open alternatives exist) as “crap” regardless of the quality – just a habit of mine
.
Clear. Just for a record: it *was* open for a while, too bad it isn’t now. I have the same attitude when it comes to closed-source things, but this is one of the few examples of good, closed source projects, so you know
Regards
This is why anyone who has used any sorts of alternative operating systems knows better than to depend on their systems being around long term, unless they are released under a true open source license.
How many times do we need to watch promising operating systems disappear once the companies behind them go away before we get wise to this?
SkyOS, BeOS, QNX, etc, etc…
(I’m sure others can extend that list greatly!)
–bornagainpenguin
Since QNX has been around for over 25 years and is a commercially proven platform with millions of installations, people tend to know and depend on it for just that reason. It’s not a experimental or research OS like Plan 9 or Minix, neither an also ran like BeOS.
If only BeOS source code was open-sourced back into 2002. Oh man… With all enormous work that Haiku team did in the past years and continues to do right now, they could be focused to improving OpenBeOS R6… If only…
Oh sweet Lord. Not in this Universe.
Solaris …
I’m sick of hearing people post the same erroneous view.
OpenSolaris IS licence under a proper open source licence. It can be forked at any time. Just because the licence isn’t Linux compatible, it doesn’t mean it isn’t true open source (as evident by the fact that FreeBSD has been able to implement ZFS)
OpenSolaris’ problem is that most if the developers are Oracle. So if Oracle cease development on OpenSolaris, there may not be enough of a community to fork the project. But that would still be true if the licence was BSD or GPL (possibly worse as many Linux developers would have ported the best features of Solaris to Linux and then left Oracles OS to die – at least at with the current licence, there’s an incentive to fork OpenSolaris).
The problem is that making it specifically incompatible with GPL they also closed the door to easy drivers port, and that bites big back as drivers are really a hard part to develop as the multitude of hardware to support keeps growing and, besides Microsoft and the Linux community, no other project seems to be able to keep up with the current pace. If Sun wanted to really make OpenSolaris popular, their decision about its license would be considered a mistake.
Strange.. Last time I checked, NetBSD was the most portable OS and supported the largest range of hardware.
Please Check Again.
My comment hadn’t got anything to do with Solaris being open-source or not.
Parent to my post mentioned:
“anyone who has used any sorts of alternative operating systems knows better than to depend on their systems being around long term, unless they are released under a true open source license.”
“SkyOS, BeOS, QNX, etc, etc…”
SkyOS was never open-source, BeOS wasn’t either, QNX don’t seem to have been “truly” open-source, and so on.
He cares about losing the whole OS and not being able to depend on it longer. He don’t care about the open-source state. Except if it’s open-source you can fork it even if the company behinds goes belly up and hence continue using it.
I haven’t talked about licensing at all.
Oracle seem to kill of OpenSolaris and eventually will lose more and more Solaris business and in the end just run it as part of their database business. As in you buy a server from Oracle with an OS which run Oracle well and Oracle DB on top. And that’s it.
Yes, so stop speaking out of your arse. I haven’t talked about licensing, I talked about Solaris. And Oracle mess it up even more. Solaris without Sun/Oracle behind it would probably die off.
Your GPL hate rather makes a point for that if it had been released under GPL if Solaris died / Oracle f–ked up then atleast some of the good parts could had ended up in Linux for people to switch to. But I don’t see where that would had made market sense for Oracle or Sun … “It’s good because then people can leave us!” …
Edited 2010-07-16 16:41 UTC
I don’t think anything of value remains.
Btrfs and utrace/systemtap are already the more “linuxy” ways of doing what was considered valuable in Solaris before (zfs/dtrace). I suppose dtrace was deeply intertwined with the os, and zfs reinvents the whole FS stack from top to bottom.
So basically we don’t need to fret about the licensing decision anymore.
You are so wrong. Wrong wrong wrong wrong wrong.
You can fork OpenSolaris. I’ve already stated this.
Did you just completely ignore my post?
Grow up.
I don’t hate GPL. Quite the opposite in fact – 90% computers I manage run Linux as the main OS and ArchLinux is my primary desktop OS. I just don’t believe that GPL the only licence worth considering. In fact I that kind of attitude is little different to OS fanboyism.
So if I come across as a GPL hater, then that’s only because you’re so blinkered to other opinions outside of GPL.
Edited 2010-07-18 21:11 UTC
You can, but no-one is likely to give a shit about it.
Get a life.
ArchLinux sucked balls the only time I tried it. Waay overhyped POS. Atleast at that date.
Archlinux is nice. But it lost the KISS distribution personality when it got infested by users who want less package dependencies at the expense of broken dependencies for example when the ‘optdependency’ thing was introduced.
run ldd on your /usr/bin and /usr/lib/ and you’ll find broken binary dependencies which can be fixed by installed ‘optdependencies’. This breaks KISS philosophy because it doesn’t confirm with how upstream designs applications.
When I tried it (To its defence it may have been pre-1.0 (0.7? 1.4? Don’t remember.)) it managed to break USB devices (switch of /DEV-system) and ALSA (probably just an upgrade / linked to old libs / whatever, the mixer stopped working) between updates.
When things break I prefer to be the one who’s f–ked up. The days then I accepted and sorta saw the fun in things breaking and “being hard to manage” has long passed by. And if it would happen I would have had enough of dealing with my own shit instead of the maintainers.
Back in those days I used to use FreeBSD but just wanted to see where Gentoo and the other “omg it’s soo good I can’t believe it’s Linux(!?
)”-distributions where at for the time being.
Things can break in the BSDs to, but when it happen in the OS it’s most likely due to an upgrade and documented with a solution in the upgrade documentation. And if it happens with ports/pkgsrc you’re most likely responsible for it yourself.
With things like Solaris it’s not like Sun would have said “oh well f–k this way of doing things, we throw it out and replace it with something else for now and let things break until we’ve fixed it all!”
And well, I think I prefer that approach. But Linux development in general seem to always has preferred new ideas, fast implementations and testing/fine tuning over time until you get the desired result over making something which work as it should at the cost of having to wait until it’s ready / a more slow and steady approach to adding new functionality.
And as far as total development tempo goes the Linux method seem to be the better one.
Edited 2010-07-20 15:46 UTC
QNX was open source only by the definition that you could see the source and check it out but it was so restricted that it wasn’t worth the hassle.
As for the community in the foundry, well there one or two active people.
It seemed more like a PR stunt.
Exactly, it was never open source by the most widely known definition, that is, OSI open source. I think that calling “open source” what is more often called “shared source” is just creating confusion. I remember the RiscOS guys did the same.
Actually, it’s much preferable for the source to be closed, because with “shared source” there’s more risk they can wrongly or falsely accuse someone of copyright infringement, whereas that’s not the case when the source is not available.
This question might be slightly off topic, but is QNX the OS that could fit on a floppy disk back in the day and have a full GUI? I think it was around 2001 – 2002 a friend of mine was showing me that OS and I could swear it was QNX. Maybe it was Q-something else?
This is the same company, but that demo floppy was QNX4 not QNX6.. sadly they didn’t do enough to market the OS for desktop/workstation usage.
QNX has been around for a very long time, used in a variety of embedded devices.
In Ontario, Canada, QNX2 was used on the Unisys Icon family of computers.. used to be entire computer labs full of them, from the 1980’s to the mid 90’s anyway.
Edited 2010-07-16 20:51 UTC
Yup, the QNX demo disk. Worked quite well on the Compaq Armada 7800 I owned at the time.
More info, incl. screenshots here:
http://toastytech.com/guis/qnxdemo.html