3 ms·
> QNX was cautiously courting the open source concept and venturing in the direction of shared source (with some code already available), when BlackBerry bought
by i336_ 9y ago
> QNX was cautiously courting the open source concept and venturing in the direction of shared source (with some code already available), when BlackBerry bought them and threw all of that out the window. Biggest yanked opportunity. D:
I said that ^ last time QNX got brought up:
https://news.ycombinator.com/item?id=12777520 https://news.ycombinator.com/item?id=12777520
Quick UNix (QUnix, which became QNX "after a brief infestation of AT&T lawyers" - https://news.ycombinator.com/item?id=4834334 https://news.ycombinator.com/item?id=4834334) seems to be one of the cooler UNIX designs (and OS designs) out there. (As an aside, does anyone know anything about Dan Dodge's "hand-made modem" referenced in that thread? There are no citations I can follow, I've been curious for months.)
Sure, it probably has its fair share of little insanities (which are not common knowledge due to it being closed source) but it seems the things QNX gets right more than make up for them.
I have three very pointed questions. I'd love answers, suggestions or critique to any or all of them.
1. Does anyone know someone who knows someone [...] who might be able to unearth "the right person to talk to" at BlackBerry about QNX? While I'd probably want to defer the actual conversation to a person (or small group) more familiar with QNX from real-world experience in industry, my motivation behind adding this point is the hope that there may still be some glimmer of interest in a shared-source model, and that - and I really hope this - the source-access shutdown was simply due to insufficient time to refine the contracts (because the buyout was executed at a particularly fast-moving instant in time). My thinking is that BlackBerry doesn't realize the QNX enthusiast community exists because they were shut down, and that if BlackBerry could be made aware of the fact that there are still big fans out there, it might be a simple matter of adding a couple clauses to some super-high-level document to eg re-allow academic access. (Yes, this point has a lot of naïveté in it, but one can dream, right?)
1a. As a continuation of the above question, if I had a question/notation for BlackBerry/QNX, it would be this: a) QNX had a shared-source release, and seemed to be making good progress with that (in short, people weren't running off with the code); b) the automotive industry and other large interests sound like fairly stable considerations for QNX; could this affect (a) in the future in any way? If this were possible, would it be possible for Photon to be included (perhaps with no warranty or code updates)?
2. Can anyone who still has an old copy of the shared-source releases from before the BlackBerry buyout dig out the license files? I'm very very curious to see if there are any loopholes I could edge through sideways...?...
3. There is terribly little easily-findable documentation on the Internet helpful to OS enthusiasts who have ahem stumbled on certain TTHs and want to explore what they have found solely for personal research. While I had success with QNX 4 some time ago (note what I am saying), I'm quite stumped about what to do for any version beyond that. I'm very interested to learn whatever I can here, perhaps via email (where I can share notes too).
- jacquesm 9y agoI've gone that road, even had financing lined up for a 'ransomware' style buy-out. Nobody even wanted to talk. Very frustrating. I completely share your sentiment and your vision.
- i336_ 9y agoWow. That's quite something. For the whole OS?! :D It seems that BB bought QNX for $200M: http://blog.vdcresearch.com/embedded_sw/2010/04/update-2-rim-to-acquire-qnx-terms-of-the-deal.html http://blog.vdcresearch.com/embedded_sw/2010/04/update-2-rim... If I may ask, I'm very curious (and I suspect others may also be interested in this datapoint) - what sort of ballpark......? And what were you requesting? Also, I have to admit that I'm unsure what you mean by "'randomware' style buy-out" - the only definition of ransomware I'm finding on Google (searching for "randomware buyout", heheh) throws back info about malware. I'm not as well-versed in business to be able to intuit what you're getting at there :) You've probably noticed that I also replied to your top-level comment.
- jacquesm 9y ago> If I may ask, I'm very curious (and I suspect others may also be interested in this datapoint) - what sort of ballpark......? Well, you already supplied a number. I figured since they had not done QNX any favors they should be happy with a bit of a discount, especially because I did not care for the branding. > And what were you requesting? Open source the complete OS as it was available at that time, without the brand. Perpetual license (they'd already done a release-and-retract once before). Other deal parameters to be negotiated. Anyway, as it is I strongly suspect that QNX will go under with BB at some point, or that it will get sold to the highest bidder in some kind of bankruptcy proceedings. Which likely will be quite a bit cheaper :) I already wrote one QNX clone for 32 bits, it should not be too hard to make another for 64 bit mode and it would be a lot cheaper too so maybe that's a better way to go about it. It really isn't all that difficult, the fact that it starts off with an extremely simple kernel makes the whole problem far more tractable than it would be for a comparable macro kernel, the scope that needs inspecting to home in on a bug is a very small fraction of what you get with a macro kernel and if a device driver crashes you can simply inspect the core dump of that device driver. And with VMs as a test bed that job got so much easier.