| Weekly edition | Kernel | Security | Distributions | Search |
| Archives | Calendar | Subscribe | Write for LWN | LWN.net FAQ |
QNX source released - sort of
QNX source released - sort of
Posted Sep 12, 2007 16:06 UTC (Wed) by larryr (guest, #4030)Parent article: QNX source released - sort of
It looks to me like essential the usual "source available", "free for non-commercial use" kind of thing... essentially they seem to be going with an approach that nowadays would be considered a minimal level of "open"ness. But then again, according to the announcement, this is a move that "revolutionizes software development practices by combining the best of the open source and commercial software domains", so I must be missing something...
Larry
(Log in to post comments)
QNX source released - sort of
Posted Sep 12, 2007 16:25 UTC (Wed) by forthy (guest, #1525) [Link]
He did not say "the best of free software...". According to the "Open Source" paradigm, many eyes looking over code, and a community (people who write software for others in their spare time) are the way to success. And as marketing speaker, you must always add some "revolutionizes" and "best of" and other exaggerations, anyway.
I guess this is a half-hearted attempt to grab back some of the share embedded Linux has now. QNX probably is better in this domain (Linux is not really optimized for embedded realtime use), and adding some freeness may help. It also helps that the device makers, once they put QNX in, aren't forced to deliver sources as well.
From a Linux kernel developer point of view, people can now look into QNX sources, and judge whether it's really easier to write a monolithic or a microkernel OS.
QNX source released - sort of
Posted Sep 12, 2007 19:19 UTC (Wed) by larryr (guest, #4030) [Link]
He did not say "the best of free software...". According to the "Open Source" paradigm, many eyes looking over code, and a community (people who write software for others in their spare time) are the way to success.
I think a sine qua non of the application of the "Open Source paradigm" to software is that the software actually be Open Source software, which means that it conforms to the Open Source Definition, which includes "the license shall not restrict any party from selling or giving away the software.... The license shall not require a royalty or other fee for such sale..... The rights attached to the program must apply to all to whom the program is redistributed...."
I think it would be a specious abuse of the Open Source concept to suggest that the benefits for a software author (or anyone else) to be derived from using an Open Source license can be garnered simply by allowing people to see and talk about the source code. I think the essential distinction between Open Source and Free Software is that Open Source software is less restrictive for those to whom the software has been distributed than Free Software is.
Larry
QNX source released - sort of
Posted Sep 12, 2007 21:53 UTC (Wed) by vmole (subscriber, #111) [Link]
The distinction, such as it is, between "open source" and "free software" is that the former term was invented to avoid scaring the suits (aka "business friendly"). It was sold on the basis of "this is a way to improve your software development cycle", rather than "this is better for your users." The actual open software definition applies equally to either; it was written to ensure that users did indeed get free software; there are no extra restrictions associated with "free software". There *may* be license specific restrictions, of course. The GPL2 obviously has more restrictions than the BSD 3-clause, but both are free software licenses.
The problem is that companies have taken the term "open source" and applied to licenses that in no way pass the OSD.
They are trying to fight off Linux
Posted Sep 12, 2007 16:25 UTC (Wed) by leoc (subscriber, #39773) [Link]
Perhaps not co-incidentally, a survey recently released (discussed on linuxdevices.com here) says that embedded developers like having access to the source code for the platforms they use. IMHO the GPL is also important, but probably not as much as simply having access to the source.
They are trying to fight off Linux
Posted Sep 12, 2007 17:32 UTC (Wed) by endecotp (guest, #36428) [Link]
Absolutely. Everyone I know who has used a closed embedded OS has a horror story of spending months finding subtle bugs that would have been apparent in seconds if they had had the code. This normally involves disassembling the code and logic analysers. From an entirely pragmatic point of view, this is the greatest strength of Free/open-source software in this market segment.
They are trying to fight off Linux
Posted Sep 13, 2007 3:58 UTC (Thu) by jmorris42 (subscriber, #2203) [Link]
> Everyone I know who has used a closed embedded OS has a horror story> of spending months finding subtle bugs that would have been apparent
> in seconds if they had had the code.
Exactly the reason they are doing this and their customers will almost certainly welcome this development. Personally I think this level of openness needs to be legally mandated.
Think about it, copyright exists to "promote progress in the Sciences and Useful Arts" not to prop up a business model or produce profit. The hope for profit IS supposed to be a motivation of course, the Founding Fathers were not Communists. But this idea of keeping the source locked away where it is often lost when companies die does NOTHING to promote Progress in anything except wheel reinvention. And it is just a technical artifact of the early computer hardware, increasingly requiring more and more twisted methods to keep the idea alive. (See the PHP/Javascript code obfuscation industry.)
No, we need a better way. Copyright should be granted only on SOURCE CODE on condition that the customer get a copy of it and all the bits they need to read/compile/etc the original vendor has (except of course 3rd party closed source tools) so the code is readable and understandable by the customer. In the case of commercial code, fully protected by Copyright and not redistributable without a license, etc. But readable. You can Copyright an expression of an idea but not the Idea itself, others should be able to see it and learn. (Leaving the whole subject of software Patents for another rant.) Customers could port to new platforms, patch bugs and generally keep a product alive after being abandoned by the original vendor by distributing patchsets.
Compiled binaries should only be protected as "derived works" of the actual copyrighted work, but a court should be able to declare it void if a vendor refuses to publish the actual source that the binary was built from.