QNX "Opens" Source Code

Arista writes "QNX has announced that effective immediately, the company will open the source code to its QNX embedded, RTOS, microkernel operating system. From the press release: "Effective immediately, QNX will make source code for its award-winning, microkernel-based OS available for free download. The first source release includes the code to the QNX Neutrino microkernel, the base C library, and a variety of board support packages for popular embedded and computing hardware." OSNews features an interview with the CEO of QNX, Dan Dodge, on this announcement."

Comments

That's cool

[suso]

These are the guys that released that really cool Desktop GUI + PPP stack + web browser and OS on a single floppy disk back in the 90s. I remember also reading that the Photon GUI would let you pass applications between computers through a dock on the side of the screen. Neat stuff.

Re:That's cool

[hubert.lepicki]

Yes, these guys. But if you read not only the note that showed up on both osnews and slashdot, but the actual interview and if you download the source code - steps that I did - you will get different view ont whole "open source qnix" thing. It's "open source" only for non commercial use - which is not "open source" at all.

Re:That's cool

[porkThreeWays]

So a dual license? Blasphemy! I've never heard of a successful open source project using dual licensing!

Re:That's cool

[dougmc]

It's "open source" only for non commercial use - which is not "open source" at all.

Depends on which definition you use. Looks like the one you're using is more `free (libre) software' than `open source'.

To many, `open source' simply means the source is available. And it is.

Re:That's cool

[TheRaven64]

I've never heard of a successful open source project using dual licensing! From TFA:

We aren't releasing the OS code under an open source license. Dual licensing is fine, but none of them is open source. Also from TFA:

If fact, we're providing three licenses: one for commercial users, one for noncommercial users, and one for QNX technology partners. It's no different from Microsoft's Shared Source. None of the licenses counts as Free Software.

Re:That's cool

[crush]

Yes, it's interesting that Larry Rosen is busy spreading FUD when he claims stuff like: Open Source licenses ... relinquish[..] all control over their intellectual property and giving it away for free (pg.2) or counterpoises OSS with ... the sustainability of a royalty-based business model for commercial projects {pg.2) A lawyer involved with licensing issues should be very aware that Intellectual Property is a garbage, meaningless term usually used by people that are seeking to assert greater control over information. He should also be aware that there are very many sustainable and highly profitable Open Source .... gah, let's call them Free Software ... projects. Red Hat springs to mind as the most succesful (their entire codebase, including their build-systems and bug-reporting systems are completely Free) as do MySQL AB (ever take a look at their dual-licensing model Larry?).

Basically, this guy is being paid to say "Hey, I'm an Open Sores guy too and I say it's OK, so drink up the tasty chocolate milk and ignore the funny taste from the proprietary ingredient".

QNX is awesome, but this isn't an Free or Open Source license.

Rosen has just disgraced himself.

Re:That's cool

[abes]

Actually, it is open source software regardless. The open source refers to the fact that the source is open for all to read. It has no implications of freedom. Freedom is only guaranteed with Free Software (or sometimes FOSS). The Free of course refers to your freedom to do what you want with it.

Re:That's cool

[somersault]

Yep this was the first *nix I ever used, from a single floppy on my Amiga :D

Re:That's cool

[Mr. Slippery]

To many, `open source' simply means the source is available.

"Open source" is a term of art with a very specific meaning.

Anyone in the software field, or any related field, who thinks that "open source simply means the source is available" is dangerously ignorant.

Re:That's cool

[harlows_monkeys]

It's no different from Microsoft's Shared Source. None of the licenses counts as Free Software.

Huh? One of Microsoft's Shared Source licenses (The Permissive License) satisfies every one of the conditions RMS gives for Free Software (and every condition given by OSI for Open Source).

Re:That's cool

[MightyMartian]

When you're forced to commit the etymological fallacy to back up your argument, it's a tacit admission that your argument has collapsed.

Re:That's cool

[Nosklo]

Huh? One of Microsoft's Shared Source licenses (The Permissive License) satisfies every one of the conditions RMS gives for Free Software (and every condition given by OSI for Open Source).

But there is no useful piece of software released in this license and I doubt there ever will be.

Re:That's cool

[Hal_Porter]

And it's not possible that people disagree with your definition and use "open source" to mean "the source code is open"?

Re:That's cool

[dougmc]

"Open source" is a term of art with a very specific meaning That's one definition. Here is another. `Of or relating to source code that is available to the public'.

People redefining words to fit their agenda (for good or bad) is nothing new. And like it or not, the English language is ambiguous, and one word or phrase may mean different things to different people. And just because they use a definition that doesn't jive with the one you prefer, that doesn't mean they're `wrong'.

Anyone in the software field, or any related field, who thinks that "open source simply means the source is available" is dangerously ignorant. Anyone who speaks English but honestly thinks that words or phrases can only have one meaning is either 1) in denial or 2) doesn't really speak English.

Re:That's cool

[Hal_Porter]

It's not FUD. If they released QNX under the GPL3 they'd have allow people to pass on copies for free. So one person could buy a copy of QNX, set up a CVS server and then fork it as FreeQNX. So then potential users can then choose between paying QNX for software or downloading it for free from the FreeQNX server. What effect would that have on the price QNX can charge? Oh and they have to license their patents in a non discriminatory way too, so if they license the patents to the one person who pays, they need to license them to all the freeloaders too.

http://www.gnu.org/licenses/gpl-3.0.html

Each contributor grants you a non-exclusive, worldwide, royalty-free patent license under the contributor's essential patent claims, to make, use, sell, offer for sale, import and otherwise run, modify and propagate the contents of its contributor version.

These two things together make the GPL3 suicide for them.

Actually his objections to the GPL in an embedded world are exactly what I've been saying for ages. Especially this one -

To enable these activities, QNX intends to publish all of its runtime component source code (some source code wont be published immediately because of third-party licensing or confiden-tiality restrictions)

They can't open the source code completely, because they don't own all the rights.

And this too actually -

Technology companies implement their fundamental business strategies through licensing their intellectual property. It is a subtle task. If a company gives too much away through overly generous grants of copyrights or patents, then its competitors and customers get a "free ride" on its products; the company loses its incentive to invest in research and development.

If they followed the GPL, they don't get that return - anyone can take the software QNX paid to develop and use it for free. This is the situation that IP laws like copyrights and patents were invented to prevent - the idea is that you can invest money creating software and then license it to people because you know the law stops people who haven't paid from using it. If the software was public domain or GPLd, you can't do that - once people have the software they don't need you anymore and can decide not to pay.

You can say IP is a bad concept all you like, but it exists for a reason. In fact it exists so companies like QNX can exist.

Re:That's cool

[TheRaven64]

And this is why Free Software advocates laugh at people who say 'Open Source' is not ambiguous. The term 'open' has been used to describe a wide variety of things. Long before we had Open Source, we had OpenVMS, which was 'open' in that it completely implemented POSIX.

Re:That's cool

[crush]

It's not FUD.

Sure it is. It states that "Open Sourcing" will lead to a non-sustainable, non-commercial business. That's contention is pure FUD. I gave at least two major examples that disprove that contention. The company behind QNX would be in a prime position to derive revenue from custom development work and maintenance contracts from hardware manufacturers. Instead they're chosing to try and hold on to a confused business model where they try to fool themselves and their customers into believing that the customers have to pay for the software.

QNX has undoubted technical benefits but it's going nowhere fast with this.

Re:That's cool

[porkThreeWays]

Insightful? I was trying to be funny... =( I guess I'm gonna have to go back to prop comedy.

Re:That's cool

[ClosedSource]

It's funny that you describe the dominant and very successful business model for the software industry as "confused". Not every company wants to be in the consulting and contracts business and many customers aren't interested in paying for such services.

I have used proprietary embedded OS's before and we never would have paid for consulting services or maintenance contracts. If the OS's is of good quality and is reasonably well documented, what else would we need? However, if the OS was available for free for commercial use, we'd be happy to use it without paying for it (We wouldn't care much about having the source code: the whole point of buying it is to avoid doing the work ourselves). On the other hand, if it required us to release our code, we'd probably pass on it.

Re:That's cool

[Hal_Porter]

I explained why they can't release the software as GPL and continue to sell licenses which is what they want to. I find it amazing that people with absolutely no experience of running a software business can tell people that do that if they give their software away they can still make money from "custom development workd and maintainance contracts". I think QNX knows more about whether that's possible or not than you do.

Instead they're chosing to try and hold on to a confused business model where they try to fool themselves and their customers into believing that the customers have to pay for the software.

That's hilarious. So you say their business model is "confused" because they "try to fool themselves and their customers into believing that the customers have to pay for the software". That actually sounds like the definition of a good business model to me - sell licenses for a fee.

Whereas in your business model they stop doing that and somehow karma will make sure they still get money from other things.

Seriously, I don't know how you can say that their business model is confused without seeing the irony.

Maybe you should try the same argument next time you see something you like in a shop. Tell the assistant that their business model is confused and outdated and they should give it to you for free and then make money out of a maintainance contract. Maybe you'll get some free stuff.

Incidentally, how much money have you spent on maintainance contracts for free software? Personally, I've spent exactly $0 over my entire life. I have paid a few thousand dollars for commercial software I like though. So if I'm typical, I'd say people people who want to go on selling licensed are onto a good thing and should ignore people like you trying to trick them into giving away their stuff. But I'm sure they're smart enough to know that anyway, just like this QNX guy does.

Re:That's cool

[Anonymous Coward]

Soooo.... Mr. Lepicki, what does RMS' cock tase like?

Re:That's cool

[XNormal]

> But there is no useful piece of software released in this license and I doubt there ever will be.

IronPython.

Re:That's cool

[crush]

That actually sounds like the definition of a good business model to me - sell licenses for a fee.
Incidentally, how much money have you spent on maintainance contracts for free software?

Per annum we spend approximately $20,000 on a mix of RHEL base 12/5 and 24/7 supported systems. It's worth it. We used to spend more on VMWare, but luckily that's not happening anymore and we're using Xen.

Maybe you should try the same argument next time you see something you like in a shop.

Oh. OK. You really are confused then, maintenance contracts are not tangible goods. Software is not a tangible good. Trying to create a business model around the fiction that it's a tangible good is a dead end.

Re:That's cool

[oglueck]

I've never heard of a successful open source project using dual licensing!

You must have heard of MySQL.

Re:That's cool

[bytesex]

Exacly. BerkeleyDB is one of the very pillars of the open source industry, and it comes with exactly such a license. Don't like it ? Yield some of the power you desire and use GDBM instead.

Re:That's cool

[SiChemist]

The "IP" laws like copyright and patent were invented for one reason: "To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries;"

There are not there to encourage profit per se. They exist to encourage invention. The problem today is that, for copyright, the work essentially never reaches the public domain. Congress continues to extend the duration of copyright (under intense lobbying by "IP" firms) so that it extends, in practice, forever.

So, you are incorrect. The intent of the law was to encourage works to enter the public domain. That the current practice prevents this very thing from happening is an artifact of corporate greed, not noble intentions by the writers of the constitution.

Re:That's cool

[Mr. Slippery]

That's one definition. Here is another. `Of or relating to source code that is available to the public'.

And the American Heritage dictionary is supposed to be authoritative about software development and licensing?

Look up "trusted" in a dictionary and you won't find mention of the Orange Book or Common Criteria, but you'd better understand their definitions if you're going to talk about "trust" in a computer system.

Yes, natural language is ambiguous; one of the ways ambiguity is resolved is via context. "Work" means one thing when I'm talking about my paycheck, another if I'm talking about physics. If I said I get paid for my work on such-and-such-project and you asked how much force I exerted over what distance, you'd either be joking or you'd be dangerously confused.

"Open source", in the context of software development and licensing, is ambiguous only as a convenience for those who wish to create confusion and either sabotage, or ride the coat-tails of, the Open Source movement.

Re:That's cool

[Ruliz Galaxor]

Me is no in dineal! Englihs is no ambigurous!

Re:That's cool

[Ash Vince]

Thanks for the link, I had not read that definition of Open source before.

How does the GPL3 sit with point 6? Surely it treats different fields of endeavour differently by treating product aimed purely at commercial users and products aimed at home users differently? Doesn't it give more freedom to companies that are selling a product only aimed at other companies that it gives to companies selling a products destined for home use?

Please note that every sentence in the above paragraph ends in a question mark. I have not read the GPL3 in it entirety so cannot say for sure. Can anyone explain exactly why this does or does not apply in this case?

Re:That's cool

[Rutulian]

Uh, no. The GPL (any version) does not restrict usage at all. The only thing it requires is distribution of source code changes when distributing binaries (and in v3 anything that is required to make the source changes usable, ex: signing keys to allow it to run on hardware that requires signed binaries).

Re:That's cool

[Anonymous Coward]

You must have heard of sarcasm.

Re:That's cool

[Ash Vince]

But google base a proprietary, entirely closed product on open source technology. Here is a link:

http://www.google.com/enterprise/gsa/index.html

Why will this not fall foul of the GPL3? Surely if it contains any GPL software to make it work they need to distribute the source and any modifications necessary to make it work. I have asked this question before and someone pointed out that the Google Search Appliance was exempt as it was only aimed at commercial organisations.

I would be very grateful if someone could clear this up for me as I am just getting more confused.

Re:That's cool

[Rutulian]

Whoever told you that is wrong. If GSA uses GPL'd software (I'm not sure if it does), it has to distribute source with the binaries upon request. They don't have to give the source away for free to everyone, but they do have to give you the source changes if you purchase a binary. It also depends on how they are using the GPL'd software. If they haven't modified it and are just bundling it, they don't have to provide source for their closed application. They only have to provide source if they link against it or modify it for use in their application.

Re:That's cool

[Anonymous Coward]

How can you dispute this? If one owns a domain, say [word].com or [word].org, then one gets to define what [word] means! Are you ignorant or something?

Re:That's cool

[Anonymous Coward]

I see what you did there.

Re:That's cool

[Hal_Porter]

That's a quote from the US constitution, and while it's an an admirable document it wasn't written by economists.

Economists justify IP laws based on the idea that people who invent things should have a monopoly on using them, just like if they were tangible property. For example, suppose I invent some process to speed up hard drives and there are no patents. Unless I actually work for a company who makes hard drives I'm screwed. As soon as I publish the secret all hard drive manufacturers will use it and not pay me. The best I can manage is to sign an agreement with one hard drive manufacturer, but that's risky. The manufacturer has better lawyers than me, and I need to explain the idea before they'll talk to me. But once I explain enough of it to convince them it will work, they could just get their own engineers to fill in the gaps. Even if I manage it, the idea will end up being a trade secret and only used by the company I negotiated with.

But if I patent it I can publish the idea without putting it into the public domain - and it is this ability to publish inventions that the people that wrote the US constiution used as a justification for allowing patents. But it is not in my mind the only one. Since I work for myself, the idea that I can make money out of ideas and that those ideas are property which I can rent out to numerous third parties is the one that appeals to me because it levels the playing field when I negotiate with large companies.

If you don't have laws like that, the only way I can make money is to essentially sell hours of my time, i.e. to be a wage slave.

Re:That's cool

[josephpate]

maintenance contracts are not tangible goods. Software is not a tangible good. Trying to create a business model around the fiction that it's a tangible good is a dead end.

http://www.apple.com/itunes/store/

Re:That's cool

[Hal_Porter]

Oh. OK. You really are confused then, maintenance contracts are not tangible goods. Software is not a tangible good. Trying to create a business model around the fiction that it's a tangible good is a dead end.

So you're saying that individidual developers can only work as wage slaves? That doesn't sound much like freedom to me.

http://slashdot.org/comments.pl?sid=295451&cid=20582255

It seems like the GPL is freedom for Red Hat to use people's work and not pay for it, not freedom for the people who actually invent things to be independent of them. But they should be careful, because someone else can release their work for free or undercut them on support. Since they don't own what they're selling, they can't protect themselves against that.

Re:That's cool

[bit01]

... very successful business model ...

As in a very small number of people make a very large sum of money while everybody else is screwed? Success is in the eye of the beholder.

---

It's wrong that an intellectual property creator should not be rewarded for their work.
It's equally wrong that an IP creator should be rewarded too many times for the one piece of work, for exactly the same reasons.
Reform IP law and stop the M$/RIAA abuse.

Re:That's cool

[abes]

Firstly, I was only pointing out that labels that exist already. Open source software does not make any guarantees about ones freedom. It has no idealogical pinings, but rather simply describes having the source code available for perusale. If you don't beleive me, take a few seconds to google it.

Free software, as it's name describes, does try to make some guarantee about your freedoms in using it. There are may FOSS licenses out there, each with a slightly different idealogical bent. For example, the LGPL is less restrictive than the GPL, and is often used in libraries rather than applications.

Secondly, it's a common mistake that people make as the lines between Open Source and Free Software are often blurred.

So .. where did my argument collapse?

Re:That's cool

[ClosedSource]

"As in a very small number of people make a very large sum of money while everybody else is screwed?"

I think you have a basic misunderstanding of the software market. Some companies make a lot of money while others only get modest return on their investment. As in any business, there are many more consumers in the marketplace than there are producers.

"It's wrong that an intellectual property creator should not be rewarded for their work.
It's equally wrong that an IP creator should be rewarded too many times for the one piece of work, for exactly the same reasons."

What are those reasons that you claim are exactly the same in both cases. Even if I agreed with your claim (which I don't) I'd say the reasons would actually be quite different. But the main point is who decides when an IP creator has been rewarded "too many times"?

I could make a similar unsupported claim about services "It's wrong that a company providing services should be rewarded too many times for the one body of knowledge they possess." In other words, it's OK for Red Hat to charge a single customer (or pick your favorite number here) for updates, but since that update is "one piece of work", additional copies should be available at no charge.

I think that a consistent application of this "too many times" theory will render most businesses "wrong".

Re:That's cool

[Hooya]

> Anyone who speaks English but honestly thinks that words or phrases can only have one meaning is either 1) in denial or 2) doesn't really speak English.

or 3) understands English all too well and therefore has a whole different interpretation of 'honestly'.

Re:That's cool

[mwvdlee]

So distribution of the code is allowed as long as it's non-commercial?

And are companies allowed to use such a third party distribution commercially if they pay QNX license fees?

Seems to me the only thing they are limiting is free (as in beer) commercial use. It probably doesn't fit the OSI definition, but it's useful nonetheless.

Re:That's cool

[Ash Vince]

Go and read the link here:

http://www.google.com/support/gsa/bin/answer.py?answer=15898

It says that the google search applicance uses a version of Linux.

Re:That's cool

[adah]

I've never heard of a successful open source project using dual licensing!

So you do not use Vim? Damned evil member of the Church of Emacs.

Re:That's cool

[anothy]

"Open source", in the context of software development and licensing, is ambiguous only as a convenience for those who wish to create confusion and either sabotage, or ride the coat-tails of, the Open Source movement.

the "Open Source movement" was started in, what, 1997? it was an attempt at formalizing an existing ethic of open source software (read: software who's source code you could read). OSI is free to use "Open Source" as a trademark, but they do not thereby get the rights to redefine common english words with pre-existing domain-specific meanings. the term "open system" was in the wild at least as early as 1980, and "open source", as commonly used before 1998, is a parallel concept.

nobody here's disputing the OSI's rights to their trademark, but the fact is they chose a term (chosen in a single day "strategy session", by an engineer, described by the OSI as "the best they could come up with") with a pre-existing meaning in the context they were concerned with. it's worth noting that OSI themselves claim no (legal) right on marking things "Open Source(TM)"; if you want that meaning, you mean "OSI Approved" or similar.

Re:That's cool

[Anonymous Coward]

Dude... this is slashdot. You are using critical thinking. Critical thinking must be checked at the door. We'll let it by this time...

Re:That's cool

[Nosklo]

> But there is no useful piece of software released in this license and I doubt there ever will be.
IronPython.

Useful? Useful to them, because the only place I can see IronPython being useful is in a very microsoft-tied project. Well-planned projects would use CPython instead.

Let me be the first to say....

[Corporate Troll]

Cool!

Voting machines

[goombah99]

This is huge news. One of the most popular paper ballot systems, the ES&S model 100 optical scan runs on QNX. this means it is now theoretically possible that ES&S could go open source if they wanted to.

Re:Voting machines

[squiggleslash]

Not unless they switch from QNX. QNX is not going open source, it's a shared source scheme (specifically, commercial use requires payment.)

Re:Voting machines

[Chyeld]

They are "Opening the source" but not going open source. Basically if you are non-commercial (i.e. academic or amateur) you can get a free license to play around with the code, but if you are commercial then you have to pay.

This is more like one of Microsoft's aborted attempts at opening source, where they'll let you have access as long as you don't do anything they might have wanted to do or competes in anyway with them.

Re:Voting machines

[friedman101]

They already did, here's a snippet

int maxDonation=0;
int bestCandidate=0;
for (int a =0; a<= numCandidates-1; a++) {
if (candidates[a].donation > maxDonation) {
maxDonation = candidates[a].donation;
bestCandidate = a;
}
}

candidates[a].CastVote();

Re:Voting machines

[nacturation]

for (int a =0; a<= numCandidates-1; a++) { Interesting... most people would just write:

  for (int a=0; a < numCandidates; a++) {

but I guess that's what makes the code so funny.

Re:Voting machines

[friedman101]

i certainly hope most people wouldn't write it that way!

if you have one candidate then numCandidates=1

you'd exceed the maximum index of the array (which begins at zero) if you didn't subtract one

Re:Voting machines

[Anonymous Coward]

The two statements are identical in operation, the fact he's using "less than" rather than "less than or equal to" means he doesn't need the "minus one".

Re:Voting machines

[rar]

int maxDonation=0;
int bestCandidate=0;
for (int a =0; a<= numCandidates-1; a++) {
if (candidates[a].donation > maxDonation) {
maxDonation = candidates[a].donation;
bestCandidate = a;
}
}
 
candidates[a].CastVote();

Not only corrupt, but also buggy. It always casts the vote for the last guy + 1, overflowing the candidates array. Apparently the last defense of democracy is that people code like crap.

Re:Voting machines

[ZerothAngel]

Not if you use less-than (<) as nacturation's post suggested...

Re:Voting machines

[TheRaven64]

No, because you start at 0, then you add 1, and then check if 1 1, which fails, so you only execute once. In the original, with the -1, if numCandidates is 0, then you get an integer overflow. Since you are using signed integers (why? The negative values are invalid for these quantities) then this will not be a problem, but if you were using unsigned ints then numCandidates - 1 would be (2^n)-1 (where n is 32 or 64, or occasionally 16), making this a very dangerous practice since it has the potential to loop thousands or millions of times, writing over undefined memory.

Re:Voting machines

[dslauson]

"Not only corrupt, but also buggy. It always casts the vote for the last guy + 1, overflowing the candidates array."

Since we're nitpicking, it actually won't compile at all because the variable "a" has gone out of scope on the last line.

God, I hate myself sometimes.

Re:Voting machines

[goombah99]

Ah thank you for reminding me why I stopped programming in C.

Re:Voting machines

[Zebedeu]

Isn't the variable 'a' defined only inside the for loop?
It seems to me that the code wouldn't even compile, but I guess that it might depend on the language used...

Re:Voting machines

[boa]

AFAICT, the loop stops when a == numcandidates - 1. Assuming that the candidates array is zero based and has numCandidates elements, the array access should be ok.

The coding style "i http://www.thedailywtf.com/ ;-)

Re:Voting machines

[ConceptJunkie]

But the code is using signed integers.

Of course, I agree with you that the -1 version is suboptimal from a clarity point of view.

Thank you for reminding me why I love programming in C(++): Because you always know exactly what will happen, and its potential effects.

Re:Voting machines

[Paradise Pete]

it actually won't compile at all because the variable "a" has gone out of scope on the last line.

Which would be lucky for him, since then he has a chance to realize that the variable he wants to use in that line is bestCandidate, not a.

Re:Voting machines

[Anonymous Coward]

They already did, here's a snippet

It's amazing to me how this type of posted code snippet is almost always just horrible code. Does everybody write code like that? I can just imagine armies of corporate coders spending 40% of their time fixing compile-time errors, another 58% figuring out why their code doesn't do what they think it does, and then just 2% creating new code.

In this example, when he finally figures out how to get it to compile, with his probable "solution" to declare a outside of the loop because "I don't know why - that's the only way to get it to work," he'll then have no idea why candidate[numCandidates] gets all the votes. And when he somehow eventually manages to make it work he still won't know exactly why. It's a miracle any program ever gets finished.

Re:Voting machines

[Anonymous Coward]

It always casts the vote for the last guy + 1, overflowing the candidates array.

-- thereby electing Ron Paul.

Re:Voting machines

[noidentity]

Nope, the loop stops at a > numCandidates-1, that is, a = numCandidates, which is past the end of the array.

Not sure what's wrong with the coding style. Restoring the whitespace yields

int maxDonation=0;
int bestCandidate=0;
for (int a =0; a<= numCandidates-1; a++)
{
    if (candidates[a].donation > maxDonation)
    {
        maxDonation = candidates[a].donation;
        bestCandidate = a;
    }
}
 
candidates[a].CastVote();

About all I'd change is the last line to use bestCandidate, and the for loop to the more common "a lt; numCandidates" (I'm guessing that the original poster simply doesn't regularly code in C++/Java).

Re:Voting machines

[boa]

You're correct regarding the loop. How could I get that wrong? :-( Regarding the coding style, slashdot ate most of my comment as it contained a couple of < . What I *tried* to say is that it is silly to loop using i <= foo - 1 and much better to use i < foo.

Microkernel? WTF?!

[crush]

Don't they know that it's standard wisdom on Slashdot that microkernels can't work? What's wrong with these guys?!!! Myself, I'm still waiting for GNU/Hurd :)

Licenses?

[Anonymous Coward]

Any chance of finding out which 3 licences you can choose from without regestering to yet another website?

Re:Microkernel? WTF?!

[Anonymous Coward]

QNX is real time isn't it? That's supposed to be the one application where you want a microkernel.

And it's not that they can't work, it's just for general purpose, they tend to be slow.

Excellent news

[dargaud]

QNX has some of the best real-time features of any OS and its message passing architecture is reliable and pretty simple to use. The main problem so far was its price, lack of source and overall lack of applications. This will likely change quickly if it is open-sourced. I can see it become a serious contender to the various complex and poorly documented patches to turn Linux into a real-time system. Excellent news indeed.

Re:Excellent news

[TheRaven64]

I agree. QNX has the most elegant design of any kernel I've seen. Unfortunately, it's not being open sourced. The source will be available, but not under a license compatible with the OSI's Open Source definition or the FSF's four freedoms.

Re:Excellent news

[dargaud]

I knew there must have been a catch in there. Hah, back to reading Hallinan's Embedded Linux Primer...

Re:Excellent news

[mi]

The source will be available, but not under a license compatible with the OSI's Open Source definition or the FSF's four freedoms.

I would like to know, what it is, that you (or someone you know) are currently doing with a "truly free" OS, that you will not be able to do with the QNX now.

Please, advise.

Re:Excellent news

[Anonymous Coward]

Modify the source code for commercial gain without having to pay a licensing fee for each device, which can add up to millions of dollars for any reasonably successful product. I'd much rather abide by the terms of the GPL than the terms of this QNX "open source" license.

Re:Excellent news

[hawk]

So it's free as in speech, not beer :)

hawk

Re:Excellent news

[Tyger]

If you read some of the source, the elegance is retained there. A company I used to work at had a full license to the source of QNX, including Neutrino which, at the time, was rare. Of all the licensed commercial sources I've worked with, QNX was the only one I could say was a pleasure to work with. Most feel about as pleasurable to work with as a root canal. For a project of the size, it's actually very well organized and written.

It's just a shame it's not a full open source.

Re:Excellent news

[rh0]

Uh . . .

I believe that one of the ways QNX is able to make real-time guarantees is by not using virtual memory. All applications run in memory -- so you don't have to worry about the additional overhead of swapping.

What I'm saying is that having virtual memory and offer real-time performance guarantees are kind of mutually exclusive conditions, and both probably can't exist together in the main kernel.

Just my 2 cents.

Re:Excellent news

[Anonymous Coward]

Virtual memory and memory resident are not mutually exclusive concepts.

All processes in Neutrino run in virtual, protected memory address regions. But yes, they also run memory resident.

You may be referring to paging systems that swap memory regions out to disk. That would be different. So yes, the Neutrino kernel supports virtual protected memory addressing and hard real time scheduling.

(disclosure: I am not a QNX employee, but I have designed and developed Neutrino applications for core telecom network equipment.)

In Neutrino, only the scheduler and interrupt handler run in kernel space, all else runs in virtual mapped user land. So if a device driver has a severe defect, that driver can crash (and dump core) without bringing down the kernel. This allows far easier debugging. When you write a driver, the trick is to provide as minimal interrupt processing as possible (since your interrupt callout runs in the kernel) and delegate the bulk of the processing to the user land interrupt handler. It seems unnatural until you have written a couple of drivers, but once you have, it is pure joy.

Another important concept is the use of thread pools within drivers to properly support the Neutrino priority inheritance model. If you are not using thread pools in your driver, you are not handling priorities properly, and you are probably going to encounter resource scheduling issues eventually.

Re:Excellent news

[rh0]

Yeah, I guess I stuck my foot in my mouth, and botched my terminology.

Technically speaking, QNX does use virtual memory -- however, it doesn't support virtual paging (if I understand it correctly). So pages can't be swapped into and out of memory, and all apps must be memory resident (as you've pointed out).

I believe there are add-ons (potentially third party?) implemented as servers which will support swapping to disk. But these add-ons jeopardize the real-time performance of the microkernel.

So, if I can correct myself, I do believe that supporting virtual pages in secondary storage and being real-time are probably mutually exclusive conditions.

Feel free to correct me if you disagree.

Re:Excellent news

[mi]

Modify the source code for commercial gain without having to pay a licensing fee for each device

Uhm, yes, if you are after commercial gain, then QNX wants a slice. Nothing wrong with that.

But my question was, what are you (or someone you know) currently doing, that you would not be able to do with QNX?..

In other words, I don't believe, a single Slashdot user is currently selling a modified Linux/BSD code and would like to do the same with QNX, but is prevented from doing so by the latter's restrictive licensing... All of the bitching is by people, who aren't selling anything anyway (if they were, they would not have minded paying QNX their licensing fee).

Re:Excellent news

[bit01]

... who aren't selling anything anyway (if they were, they would not have minded paying QNX their licensing fee).

Nonsense. Want to maximize profit? Use the OS with the lowest licensing fee. i.e. not QNX.

---

Beware deceptive astroturfers.

Re:Excellent news

[mi]

Nonsense. Want to maximize profit? Use the OS with the lowest licensing fee. i.e. not QNX.

You are changing the subject. I was looking for anyone, who is currently doing something with a properly free OS, and who will not be doing the same with QNX due to the latter's more restrictive licensing. If you don't have such an example, you should not have followed up.

You especially should not have followed up, because you posted nonsense — your statement would only be true, if the OSes being compared were otherwise equivalent. They aren't. Each has strengths and weaknesses, and for many paying QNX would be (far) less expensive, than keeping extra developers on stuff to deal with a free OS' idiosyncrasies.

Do try harder the next time around...

Re:Excellent news

[bit01]

Nonsense. Want to maximize profit? Use the OS with the lowest licensing fee. i.e. not QNX.

You are changing the subject. I was looking for anyone, who is currently doing something with a properly free OS, and who will not be doing the same with QNX due to the latter's more restrictive licensing. If you don't have such an example, you should not have followed up.

Okay, I'll spell it out: Anybody selling a large volume of something where the marginal cost of tweaking a free OS can be amortized over a large number of units. e.g. Tivo. Or organizations with more staff time than money e.g. Beowulf cluster. Or where they want to distribute the device for free to third parties including commercial e.g. OLPC. Or they want to do their own maintenance and modifications and sell the result. e.g. RedHat. Or a large company with deep pockets that doesn't to be beholden to a small company that might raise the price on them once they're committed. e.g. IBM

You especially should not have followed up, because you posted nonsense -- your statement would only be true, if the OSes being compared were otherwise equivalent. They aren't. Each has strengths and weaknesses, and for many paying QNX would be (far) less expensive, than keeping extra developers on stuff to deal with a free OS' idiosyncrasies.

Given the power and cost of hardware these days QNX and Linux are equivalent for almost all applications. The QNX company is probably feeling some pain because of this. QNX does have a niche but I suspect it's getting smaller all the time, particularly since QNX can't leverage the user/development base that linux and other free software can.

Don't get me wrong, I like QNX from a technical perspective but from the point of view of establishing a standard they blew it by overpricing and it's too late to change now. A company I worked for considered it once but rejected it at the time because of cost, lack of third party tools and various license restrictions that impacted our product.

Do try harder the next time around...

Please be more polite next time.

---

Any large public or private organisation paying recurring, per-seat licensing for software is being economically stupid.

Re:Excellent news

[marcosdumay]

No, it won't be free on any of those senses. It will be "open" tough, so every customer will be able to look at the code (and probably fix bugs).

Re:Excellent news

[larytet]

if the license language follows the one of WinCE (MS) I would not touch it.

WinCE will cost you $800 and You have to call a representative and prove that you are a major manufacturer or an university or something to get the source code.

If the licenses is not GPL or BSD or any other "free" open source license they did about nothing. Indeed ability to look into source code is interesting, but in reality the need arises rarely. I spent 10+ years in vxWorks world without access to any source code and applications I developed work just fine. The same with WinCE - I had an access to part of the code, but frankly I never looked there.

I can try to put a short list of reasons why You can ask for the OS source code - you are a kernel developer porting the OS to new CPU or a board radically different from the existing and supported already - you are a manufacturer who strips (or inflates) the OS to fit the marketing needs and discovers that kernel's API is not flexible enough for yet another 10 drivers. - you need every microseconds of the CPU performance and literally count CPU cycles and you need patches in the scheduling scheme or TCP/IP or packet filter - you work with relatively large and/or young and/or buggy and/or bad documented kernel and unstable (backward compatibility) kernel API (i love Linux btw, but the truth is ... you know ... )

The list above is not full, but probably covers some of the possible reasons. How QNX fits here ? I bet they provided source code to their large customers. These news is just another sign, that competition with Linux in embedded world heats up. IMHO QNX is may be 5 years late.

Under what license

[kensai]

I didn't see it in the article, but what license are they releasing this under I wonder. Will be one of the standard ones or will it be YAOSL (Yet another open source license)?

Re:Under what license

[betterunixthanunix]

TFA:

" under a new hybrid software licensing arrangement. "

And:

" Access to QNX source code is free, but commercial deployments of QNX Neutrino runtime components still require royalties, and commercial developers will continue to pay for QNX Momentics® development seats. "

(Hint: It's definitely not GPL)

Re:Under what license

[Anonymous Coward]

Here's an article that explain the license and model

http://www.qnx.com/download/feature.html?programid=16868

and an FAQ:

http://community.qnx.com/sf/wiki/do/viewPage/projects.community/wiki/FAQ

Re:Under what license

[Anonymous Coward]

From the interview: "The initial release includes the source code for the QNX Neutrino microkernel - the crown jewels of our OS technology. The release also provides source code for many board support packages, published under the Apache 2.0 license."

License key?

[Constantine XVI]

I've been trying to get the non-commercial key from them for Momentics, but it doesn't seem like they plan on sending it to me anytime soon. Anyone else having any luck?

Re:License key?

[Anonymous Coward]

Each time I've needed to get a key it's been easier. Last time I had to get a purchase order (for $0). If you're not part of an academic institution, you might have trouble.

Re:License key?

[nurb432]

If you use the webpage and choose the 'personal' license, it shows up in your email in about 20 minutes.

No fuss at all.

Under what licence?

[Ant P.]

I had a look on the linked site and I don't see any mention of what they've released this under. I tried to see if it was in a file in the source tree, but... I get a "permission denied" message going there.

Re:Under what licence?

[g4b]

Without knowing the license type, well could mean everything.

Isn't Java Opensource, too?

However for learning purposes it's still great news. QNX is quite something even people happen to know who can't really tell the difference between Linux and BSD...

I bet it isn't GPLv3 d'oh

That's cool-Tivo.

[Anonymous Coward]

"It's "open source" only for non commercial use - which is not "open source" at all."

Well considering the changes in the GPL and all the talk here. I'd say that that's plenty open enough.

Re:That's cool-Tivo.

[cortana]

Please show me where the new GPL prohibits commercial use of covered works?

Re:Microkernel? WTF?!

[Spy der Mann]

Myself, I'm still waiting for GNU/Hurd

Oh right, I heard Duke Nukem forever requires it.

Danh

[Anonymous Coward]

Dan Hildebrand is rolling in his grave.

Is QNX dead?

[Anonymous Coward]

I guess it has no more value if they are going to give it away. It's dead in the sense that OS/2 and CP/M are dead.

Re:Is QNX dead?

[Just Some Guy]

It's dead in the sense that OS/2 and CP/M are dead.

Actually, I think it's dead in the sense that Solaris is dead, meaning that it ain't.

Re:Is QNX dead?

[f8l_0e]

Quite right. OS/2 still lives on even if under a new name. http://www.ecomstation.com/

Re:Microkernel? WTF?!

[LurkerXXX]

BeOS had a microkernel, and it was not slow.

Source Available, NOT Open Source

[Russ Nelson]

Sheesh, this is no better than Microsoft's "Shared Source"! They restrict commercial development, just like Microsoft.

This is Source Available software, NOT Open Source Software. You don't have all the freedoms available to you that are described by the Open Source Definition.

Re:Source Available, NOT Open Source

[Anonymous Coward]

oh shut up you fucking fundamentalist commie. this is way better than not having access to the source code at all.

Re:Source Available, NOT Open Source

[Anonymous Coward]

Why is this marked flamebait? It's more insightful than the parent post.

Re:Source Available, NOT Open Source

[Bloodoflethe]

I believe you have confused the Open Source Definition, based on the Free Software Definition. Commercial Development is not restricted. Free development is still allowed (Hint: free does not refer to price, but rather freedom).

Re:Source Available, NOT Open Source

[fast penguin]

As far as I understood it, you are free to hack on QNX, it just isn't royalty-free. If you make changes, and want to distribute them, you have to do so in the form of patches, and the recipient would of course need a copy of QNX. I can see QNX buying patches from the community to merge with the mainline. Could be a successful business model, and still close to the openness of free software (just not as in free beer).

Microsoft shared source is a difference beast. You couldn't compile the thing, you could just had a look at some of the components, it wasn't hackable. It was also damn expensive, while QNX source code comes along with a software license.

Re:Source Available, NOT Open Source

[Russ Nelson]

You're not free to distribute your code commercially. Even RMS believe that commercial distribution is okay. He disagrees with *proprietary* distribution, sure, but he's fine with commercial distribution. QNX isn't -- and that's why the code is neither Free Software nor Open Source.

Re:Source Available, NOT Open Source

[Russ Nelson]

you are free to hack on QNX, it just isn't royalty-free.

Royalty-free redistribution is one of the freedoms that you get with Open Source. Free Software, too.

Without it, you cut your community in half: non-profit, and for-profit.

Re:Source Available, NOT Open Source

[kernelpanicked]

Aye, occasionally an AC does post something worthy after all. I have really had it as well with the "It's not GPL, so it's not really open source" nonsense. The fact that the source is now free for me to pull down and hack on as I please is good enough. If I decide to make a commercial venture with the changes I've made, I see no problem with paying back royalties on the hard work that others put into the OS to begin with.

Re:Source Available, NOT Open Source

[Russ Nelson]

Oh, it *is* a good thing. But it's not Open Source. Named things have characteristics. A vegetarian is someone who doesn't eat flesh. A pacifist is someone who won't pick up a gun to solve a dispute. If they do, then they aren't. And if you prohibit royalty-free distribution, you're not using Open Source.

Re:Source Available, NOT Open Source

[squiggleslash]

oh shut up you fucking fundamentalist commie

That's pretty funny given who you're replying to.

this is way better than not having access to the source code at all.

Actually, it may not be. One of the problems with non-Free source is that there's the risk of code tainting. Someone who's exposed to encumbered source code may later find it harder to prove they didn't infringe copyrights if they've released code that's similar.

This was one of the complaints levied against Sun's Java policy back when it was shared source and Sun had to jump through several hoops to reassure everyone on the subject. It may well have been one of the issues that Sun considered important when they finally relented and went GPL.

The best that can be said for this is that it's possible it'll act as a thin end of a wedge for QNX that means they'll eventually realize splitting the community into commercial and non-commercial groups, and having their code untouched by communities worried about the risk of tainting, is bad and they'll go the whole way, adopting a good copyleft for those willing to share their changes, and an alternative license - with money changing hands - for those unwilling (as TrollTech has with QT, and Sun has with Java.)

Sun as the company with the biggest example of such a project needs to show the way here.

Re:Source Available, NOT Open Source

[MBGMorden]

You can't prop up your argument based on definitions while trying to redefine the term at hand. Open source means, wait for it, the source is open. Royalty-free distribution doesn't have jack squat to do with whether the source is open or not.

Re:Source Available, NOT Open Source

[CowboyBob500]

We've had this discussion before, and the OSI have *absolutely* no authority to claim any ownership on the term or definition open source. Open source to me (and most other people) means that the source is available for examination. End of story.

I'm beginning to get to the end of my tether with the zealotry and politics involved in the FOSS community. It's rapidly becoming a cult where no sense can be talked to the brainwashed members.

Bob

Re:Source Available, NOT Open Source

[Anonymous Coward]

Probably from the "oh shut up you fucking fundamentalist commie". Just a guess.

Re:Source Available, NOT Open Source

[Russ Nelson]

Errrr, except for Rip Van Winkle, a few time travellers who skipped over the last ten years, and a few trolls out there who are deliberately obtuse, everybody knows that Open Source means the Open Source Definition.

Re:Source Available, NOT Open Source

[Russ Nelson]

Open source to me (and most other people) means that the source is available for examination. End of story.

Ahhhh, you must be one of the deliberately obtuse trolls to whom I refer in my previous post. If not, then show me your beard or time machine. Attempting to claim that "Open Source" sprang from your forehead fully formed without any context is simply not plausible. Sure, you can find a few people who used "Open Source" before 1998. But if you want to claim that OSI had nothing to do with the popularity of "Open Source" after then, you have to rewrite history. Here's the wikipedia link, good luck, start editing!

Re:Source Available, NOT Open Source

[w000t]

Are you trolling or just retarded? Nobody is claiming that the code won't be made available for everyone to see, just that when something is labeled as "Open Source", people who reads this site (by a large number, despite your claims) usually expect another whole other bunch of things (rights) with it (damn, even most anti-F/OSS people understand that much).

I'm beginning to get to the end of my tether with the zealotry and politics involved in the FOSS community. It's rapidly becoming a cult where no sense can be talked to the brainwashed members.

You blindly refusing to understand the discussion doesn't make F/OSS proponents irrational or brainwashed. And you having a similar argument with someone some other time doesn't make you right either. You're the one pushing the narrow view you're bitching about on others ("End of story"?).
For the record, I agree that the OSI doesn't have any more rights than you and my to define what open source is. They just happen to provide the definition most people here use.

Re:Source Available, NOT Open Source

[CowboyBob500]

No I'm not a troll and I understand the history. Just that I don't recognise the authority of the OSI to define such terms. I've been through this on another thread with Bruce Perens here and he couldn't come up with any valid reasons that I (or others) could see for claiming that authority.

Bob

Re:Source Available, NOT Open Source

[CowboyBob500]

Are you trolling or just retarded?

Thanks for vindicating my statement about zealotry for me...

They just happen to provide the definition most people here use.

I'm talking about the definition the world at large uses, not the definition of a bunch of people on a tech blog.

Bob

Re:Source Available, NOT Open Source

[squiggleslash]

Perens wasn't claiming any "authority" in the sense you were describing, which is why you can't get the answers you want out of him. He was describing what Open Source is, based upon community consensus (ie, that it's the same thing in practical terms as Free Software, but described with a development model bias instead of freedom bias.) This is a little like you claiming a car is a grapefruit, someone else saying "No, look at this dictionary, it defines it as a small motorized vehicle with four wheels", and you claiming the authors of the dictionary do not have the authority to define it as that. That's irrelevent! They've done the research, they know what it means, they're not lying when they give it that definition. They have the authority to say that's what it commonly means, just as Bruce does.

While you continue to ask questions like this, you'll never get useful answers. Open Source is what it is. Most people, in the real world, who know enough about computing to understand concepts like "source code", define Open Source the same way Russ Nelson, Bruce Perens, ESR, Michael Tiemann, Linus Torvalds, RedHat, Canonical, IBM, Sun Microsystems, Microsoft, Apple Computer, the BBC, Theo De Raadt, the majority of people on Slashdot, Hans Reiser, my boss, various governmental and intergovernmental agencies, and I define it, and we expect it to mean that when we read the phrase. That makes it the definition. That means Bruce Perens' definition is valid. We're not being unreasonable, we're not redefining a term that was in heavy use at the time ("open source" existed, as various Google Groups searches can verify, but it was rarely used and not in any sense that contradicts the OSI definition.) And it means people who use it to mean "source available, but under a restrictive license" are misleading the majority.

We have a clear hierarchy here, and it's worth following it to be fair to the majority. There's source available, and that is made up of open source/free software and shared source. These are not unreasonable definitions, and someone who complains they are because someone had the audacity to describe open source, which somehow means they're trying to impose a definition on the majority, isn't helping.

Re:Source Available, NOT Open Source

[anothy]

Royalty-free redistribution is one of the freedoms that you get with Open Source. Free Software, too.

so says the OSI and FSF, anyway. but OSI doesn't own the words "open source". if you'd like to be more specific, use "OSI Approved" or similar; otherwise, deal with the linguistic ambiguity esr and friend introduced when they chose to name their movement and organization using a term with pre-existing industry meaning.

Re:Source Available, NOT Open Source

[anothy]

Named things have characteristics.

absolutely. but someone should've pointed that out to esr and friends before they chose to name their "movement" and resulting organization after an industry term with pre-existing meaning in the context they were concerned with. the old USENIX tape swaps were excellent examples of a widespread practice. OSI did an effective job with marketing during the Internet's pubescent growth spurt, leading some (like yourself) to believe they coined the term and defined its meaning, but the reality is that they re-used and re-defined a term with an existing use.

if esr and friends are concerned with making things clear when they're talking about code which preserves the rights they care about (which is a positive thing), they should use something without that pre-existing ambiguity. oh, wait, they already do: OSI doesn't certify things "Open Source", they certify them using "OSI Approved" and similar marks. they should stress that more; that term is clearly unambiguous (not to mention legally protected).

Re:Source Available, NOT Open Source

[w000t]

Thanks for vindicating my statement about zealotry for me... It's funny for you to say that, since I wasn't advocating F/OSS and you were the one gratuitously calling names on its proponents.

I'm talking about the definition the world at large uses, not the definition of a bunch of people on a tech blog. Yes, of course you are, and that's the whole point. You're intentionally making this argument about something it's not: an argument about what "open source" definition is the correct one, instead of about what does this QNX announcement means (something which incidentally has confused a lot of people commenting here). BTW, the OSI open source definition is not used just by "a bunch of people on a tech blog" but by almost everyone in the IT world (you know, people actually concerned with the term).

Re:Source Available, NOT Open Source

[Russ Nelson]

Nobody can come up with any valid reasons that you see, because you have your eyes shut. Nobody, except a few obtuse people, thinks "Open Source" means that your source code is open for public inspection but that nobody can make changes or redistribute it. There is just SO MUCH reference to the Open Source Definition in parallel with use of Open Source. Words exist for communication. You can use them with any meaning you want, but you fail to communicate when you use them in ways your listener doesn't expect.

They're being demolished by linux

[xtal]

I've done a few embedded linux projects over the years - we would have loved to run QNX, as I was exposed to it in university and enjoyed it, very robust, supported etc - but the licensing fees are killer. The offered advantages, at least in the applications we've worked in make it a no brainer to go embedded linux.


Access to QNX source code is free, but commercial deployments of QNX Neutrino runtime components still require royalties, and commercial developers will continue to pay for QNX Momentics® development seats.

Looks like I'll be keeping my investment in embedded linux environments. Royalty vs. no royalty with same functionality, I'll tell you who wins every time. Linux keeps getting better, too.

Re:They're being demolished by linux

[Anonymous Coward]

This is not for kids... Its not for your microcontroller little toy. This is for military environment, satellites, real stuff. Not for kids play.

And taco asking if people use it... gosh, if you dont know what you're talking about, just dont say anything.

This is industry grade, not your average webserver and stuff...

Re:They're being demolished by linux

[rastom]

Yeah, we looked at QNX and it has a very sophisticated debug interface, but one of our prime requirements was to be able to handle wireless cards, which seems to pretty much mean Linux (and quite possibly using ndis wrapper onto a Windows binary... (although that woudln't work on the more obscure archotecture we were looking at).

Re:They're being demolished by linux

[Discoflamingo13]

Which distribution did you do embedded Linux development with? LynuxWorks is the only distribution I know of which provides a comparable embedded systems OS to QNX, Green Hills, vxWorks, et al. These are all hard real-time, safety-critical OSes with certification packages for deployment in a regulated environment.

Re:They're being demolished by linux

[larytet]

yes, same here. You need that letters and faxes to send just to get an evaluation version for a period of month. Common - we are tired of this shit. I want to click and download the image and I do NOT want to send an e-mail and ask for the license key just to compile the image. and I do NOT want to look for the license file and I do NOT want my compiler to spend any CPU cycles for checking if my (floating) license is valid. And I DO want to be able to replace NIC in my server .... etc etc. This is ridiculous and annoying.

Guys, tell me that the installation is free and a simple download of TAR file and I will go to your WEB site. And, please, make the registration optional.

Good news.

[apodyopsis]

Its good news for me. I played around with QNX (around v6.1 I think) and was really impressed by the speed it could run on older machines. If I had a single CD install that could browse files, play MPS and run firefix then I would install in without a seconds thought. I remember being vaguely unimpressed with the license and installing Mandrake instead (hey, it was a few years ago..)

For a fuller office experience Ubuntu would win because of the application support, but for simple client Net/Music use in places like a lounge or bedroom QNX has got to be a good option.

Re:Good news.

[Locutus]

what's wrong with Damn Small Linux( DSL ) for that task? You did say CD, so is 50MB still too much for ya.

Also, head over to rPath and see what you can build yourself. Heck, I've even seen OS/2 stripped down to a 10MB partition and 8MB of RAM doing GUI networking, email and web browsing. Dump X and you could probably do close to the same with Linux though a full featured browser is going to take more than 8MB itself.

LoB

Re:Good news.

[everphilski]

Speed (specifically RAM consumption on an older machine). DSL is decent but QNX is tiny. The difference on a real old-school pentium is noticeable, and I can keep my GUI with QNX.

Remember kiddies, right tool for the job. Most of the time its Linux. Sometimes it is not.

Re:Good news.

[Fallingcow]

Damn straight.

There's a generation of hardware out there that doesn't run well with any OS beyond Win95, BeOS, or QNX.

Were I tasked with making a bunch of first-gen Pentium machines into internet/multimedia kiosks, my first action would be to try to get some licenses for BeOS (assuming the things needed to be 100% legal), and if that didn't work then I'd try QNX. BSD or some kind of optimized Linux with a light GUI would only be used if the other two didn't pan out, because they just can't compete performance-wise (especially with BeOS; they weren't kidding when they said that it was good at squeezing smooth video playback and GUI responsiveness out of the slow hardware available at the time)

Re:Good news.

[unfunk]

Agreed.
Linux is good and all, but it sucks at some things (last time I tried to record audio in Ubuntu ended in tears, for example)... that's the whole point of multi-boot systems, after all - to exploit the strengths of each OS.
Like it or not, there's things that Windows does better than Linux, OSX does better than Windows, and likely, QNX does better than Linux. I look forward to trying any forthcoming "OpenQNX" developments that I may try out on my MMBPC (Massively Multi-booting PC)

It's time to talk about "free software"... *again*

[Schraegstrichpunkt]

The title of the press release is "QNX Publishes Neutrino Source Code and Opens Development Process". Arista, on the other hand, didn't seem to mind mangling this in order to get this article posted to Slashdot.

I imagine this kind of thing might be why Bruce Perens said way back in 1999 that it's time to talk about "free software" again.

Re:Microkernel? WTF?!

[Anonymous Coward]

I don't remember seeing that, though I think I remember seeing that they have the same reslease date...

Non-free licence means minor event

[spungo]

It's a shame that QNX won't be released under a more free licence, as it really is a nice product. It has one massive advantage -- it's very small and slick -- it'll run on anything. It'll make a useful side-tool for development (most likely for compatibility testing), but it won't be able to compete with GNU/Linux at any real level if it ain't GPL'd.

Re:Non-free licence means minor event

[davonshire]

I disagree.
    Opening the source means people who could never afford to work with and learn this vital and amazing operating system, will have a real chance now. Just because it's not GPL'd doesn't mean this chance isn't a really good thing for people who want to program. Want to earn a living writing good software for something an industry needs and uses.

    If more companies would do things like this with their products. I think you would see a great deal more enhancements and improvements in peoples ability to find employment and enhance their skills.

    Take for example Blender3D. It's the little linux that could of the 3D CGI world. It has made amazing strides since it was purchased and it's source opened for the 3D and programming community. There you have OPEN Source. But still most major companies are using 3D Studio, Maya, Autodesk etc to make movies, commercials etc. None of who have a full version for hobbiests to use and build their talents on. While some may have education versions, most of them fall far shy of a full system so you learn how to do physics, and full lighting, cloth and hair for characters etc.

    Without the chance to really play with all the bells and whistles you can't fully explore what something can do for you. And while you may build small animations they are heavily watermarked with 'Made with Bladiblah 3D trial version' etc.

    I'm all for Free software and open source. But I'd never devalue the contribution of an effort like this.

Thank you for your patience.

Re:Non-free licence means minor event

[cdrguru]

Maybe QNX would like to survive as a software company rather than just a collection of hobbists.

That way, users of QNX needing hardware support for new platforms and processors will have it rather than hoping for something to come along.

I'm still waiting for WPA on Linux to be common to maybe be able to leave WEP behind. Of course, the hardware manufacturers that have limited budgets are choosing Linux and often leaving WPA support out.

Re:Microkernel? WTF?!

[LWATCDR]

I know that you are being funny but if you want to work on a FOSS microkernel system take a look at this.

Minix 3

It looks very interesting to me.

It's "shared", not "opened".

[darkonc]

Don't drink the PR Cool-Aid(tm) boys. Distinguish this PR hype from reality and call it what it really is -- "Shared Source". It's not Open Source(tm), and it's not "Free".

You need licenses to do things like release your own version, and that puts it in the same ballpark as Microsoft's shared source initiative.

Re:It's "shared", not "opened".

[anothy]

Don't drink the PR Cool-Aid(tm) boys.

if i promise not to drink QNX's, will you promise not to drink OSI's?

OSI does not own the words "open source". in an (implicit) acknowledgment of the ambiguity introduced by their poorly chosen name, they don't approve labeling things "Open Source(TM)"; indeed, there is no such trademark. rather, what you're probably looking for is "Open Source Initiative Approved" or similar. "open source" had industry meaning for a decade or two before OSI came along in 1998.

and, on an entirely unrelated topic, when did slashdot start replacing my (TM) (which i input as the proper Unicode trademark character) with (TM)? that's kinda obnoxious.

Mixed memories of QNX

[DrXym]

I used to program it when it wasn't a POSIX compliant kernel but some weird ass vaguely Unix-like thing. The compiler was terrible, the tools were terrible, the OpenLook-like UI was terrible but it was still a lot more liberating than programming Windows or DOS. You could do some very whacky things since it didn't really matter too much if you were talking to a process or filesystem on your own machine or somewhere else on a token ring. The real time and multitasking performance was generally excellent. I remember wasting a good portion of work just trying to port Unix and DOS tools like MicroEMACS over to it to make the environment more tolerable. In some ways the switch to Posix probably killed some of the charm QNX used to have - it was like trying to build a kit car without all the pieces. Nowadays it's just another *nix clone, even if it still focuses on the realtime market.

A Brilliant Open Source Move

[Discoflamingo13]

QNX has a viable business model just selling the test suite and certification package for the OS, without selling the OS package itself (although they will still make a killing licensing it for commercial use). They have an established base of customers in the embedded systems market who need to prove to regulatory bodies that the OS will do what it claims to do - and these documents would be worth every penny. This is an excellent way to take advantage of open source licensing.

Frustrating: QNX

[fm6]

I find the history of QNX very frustrating. When I first heard of it in the mid 80s, it was advertised as a simple Unix-like OS with very low hardware requirements. It was network-aware, supported distributed computing, and had a nice microkernel architecture.

But the most important thing was that it was a real OS, with the ability to multitask and to effectively isolate hardware from software. Contrast this with MS-DOS 3.0, which had only the most primitve, kludgy excuse for background processing. (Patterson knew zilch about os design when he set out to clone CP/M; it never occurred to him that OS code needed to be reentrant. And MS-DOS did a really lousy job of isolating hardware from software. Ironically, this fuckup assured lockin of the IBM-compatible/PC combination: software written for this platform was essential impossible to port to other platforms.

What was particularly tantalizing was that QNX claimed to run well even on very limited hardware — even 8088 systems were said to run robustly. And it shared some key features with CTOS an first-rate OS that was then dying off, due to its dependence on proprietary hardware.

The problem with QNX was that commercial license fees were very high; that's why I never played with it. It did become popular at universities (cheap academic licenses) and among certain kinds of embedded application developers (because of its nice feature set and minimal hardware requirements. I'm told that by the late 80s, most video stores used POS systems based on QNX.

Then MS-DOS/Windows started grabbing more and more of the market and QNX was forced to specialize. So for a long time now they've advertised themselves as a real-time operating system. And yes, their real-time features are very good — but they're just one part of a really good general-purpose OS.

Now, much too late to do me any good, there's an open-source version of QNX. I wish the QNX OSS community well, but there's just no place for it in the world I work in. Hopefully, embedded application developers will keep QNX alive. But I'll always be sad that QNX never found a following among common PC users — which it surely would have if the marketplace were driven by technical excellence instead of various sordid realities. This is one of the great lost opportunities in computing history. And should be a lesson to Linux advocates who think they can easily displace Microsoft.

Re:Frustrating: QNX

[Anonymous Coward]

I wish the QNX OSS community well, but there's just no place for it in the world I work in. Hopefully, embedded application developers will keep QNX alive. What's sad is that they must be doing this because QNX is losing to embedded linux, they don't see any way to compete, so they are trying anything they can to be relevant. Kind of like Sun.

Even penny hardware is good enough to run linux these days. Which is really sad, because linux is very practical and solid but it just isn't "cool" like QNX is. Nerds don't get wet dreams from 'imminently fair schedulers' and thousands of do_* functions -- or at least they shouldn't -- like they do from QNX or BeOS say.

Eventually we may get a cool microkernel OS, but it won't be QNX. It'll be something like jxos or singularity -- a typesafe safe kernel, running safe application code. Well, we can dream can't we.

Re:Frustrating: QNX

[ClosedSource]

"I find the history of QNX very frustrating"

I find your history of QNX very puzzling. MS-DOS and QNX were never really competitors since they were really in different markets.

Re:Frustrating: QNX

[ClosedSource]

You're right. Hardcore real-time embedded software is fading fast as an art. All the hard stuff is being done in hardware these days and resources are nearly unlimited. Now we are more integrators and customizers of other people's code. It sucks!

Re:Frustrating: QNX

[renoX]

Funny you said:
[[The problem with QNX was that commercial license fees were very high; that's why I never played with it.]]
and
[[But I'll always be sad that QNX never found a following among common PC users which it surely would have if the marketplace were driven by technical excellence instead of various sordid realities.]]

Sordid realities like price maybe?
Due! Microsoft has it right: start with a low cost OS, grabe a huge marketshare and *then* you can increase the price and getting huge profits!

Re:Frustrating: QNX

[xenocide2]

I think that's his point -- that QNX would have been a superior OS on the PC but for their pricing.

Re:Frustrating: QNX

[fm6]

Microsoft didn't grab its market share by low pricing. Microsoft didn't even set out to be an OS vendor. By pure blind luck, they ended up owning the officially supported OS for the IBM PC. (Bill Gates himself advised IBM to use Digital Research's CP/M, which was the leading micro OS at the time; only DR's unwillingness to sign the necessary NDAs prevented this.) That meant that any computer that pretended to be IBM-compatible had to run MS-DOS. That's the "sordid reality" I was referring to.

I don't know why QNX was so expensive. Perhaps they didn't understand the new mass market for computers. Or maybe they just realized that they couldn't compete head-to-head with Microsoft, and decided to go for niche markets that needed more functionality than MS-DOS could provide. Either way, Microsoft didn't triumph just because of its prices: there have always been plenty of low-cost alternatives to MS-DOS and Windows.

Re:Frustrating: QNX

[eric76]

There's a story about QNX I came across a few years ago.

It seems that some factory was using QNX to control a very important industrial robot. QNX had been installed and had run without flaw for a couple of years or more since it had been installed.

One fine November day, the consultant or contractor who handled that system among others was told that it had quit logging events a few months earlier. The consultant checked on it and found that the disk drive had failed the previous August.

Re:Frustrating: QNX

[fm6]

When they say "minimal hardware" they mean "minimal hardware"!

Re:Frustrating: QNX

[spitzak]

And MS-DOS did a really lousy job of isolating hardware from software. Ironically, this fuckup assured lockin of the IBM-compatible/PC combination: software written for this platform was essential impossible to port to other platforms.

Actually this is not really accurate. Early MSDOS did just about as well as anybody else at being platform-independent, more of the blame is on IBM, who threw together the "bios" really quickly and badly. MSDOS actually did run on several 8086 machines that were not IBM-compatable. The "bios" is not to be confused with the rom that everybody calls the "bios" today, but instead a block of software added to MSDOS for a particular machine (in effect a single driver for the whole machine). Other machines had enormously faster and more useful bios code.

The biggest problem with MSDOS was the 1-character-at-a-time API to the "bios", which was the same as CP/M. This added an enormous amount of overhead to any I/O. This was addressed in MSDOS 2.0, where they copied the most obvious parts of Unix and made a driver api that took a whole block of characters at a time. However by then it was too late.

IBM screwed it up by making the "bios" interface incapable of even rudimentary things. The text output had no escape sequences and could do nothing other than spew text to the screen like a terminal. The text input was incapable of distinguishing all the keys you could hit on the keyboard. And the "bios", especially the BIOS code in rom, was unbelievably slow. The lower level rom code managed to make an interface that needed TWO calls to print a character (one to move the cursor, the other to set the character), which is twice as bad as the stupidest possible api I could imagine anybody writing!

At the time, directly writing to the hardware was always going to be a big win (until the multi-character api was added). But IBM made directly writing to the hardware a *requirement* to get a working program, and their existing code magnified the win by probably an order of magnitude. It is possible that things would be different if the bios had been intelligently designed (would have taken about 1 day by any copetent programmer of CP/M). Can't guarantee it. But I don't really think the blame lies on Microsoft.

Re:Frustrating: QNX

[fm6]

OK, I stand corrected. The isolation problem was IBM's fault, not Microsoft's. (Another sordid fact: IBM's inferior design became a de-facto standard simply because it was IBM's.) Obviously IBM thought that they could get away with this design because they planned to migrate to more carefully thought out systems later. What they didn't grasp was that they had invented the first commodity computer and that its APIs had created a degree of lockin that not even they could fight. Hence the failure of the PS/2.

On the other hand, Microsoft did make those nasty mistakes in the design of MS-DOS. Or rather, they made the mistake of basing MS-DOS on QDOS, an "OS" designed by a guy who had no idea what an OS actually is! This mistake created problems that persisted to the very last version of DOS-based Windows, which was only EOLed last year!

And whether the mistakes were Microsoft's or IBM's they created a degree of lock-in that made Bill the richest man on the planet. Not superior technology, not even undersellng the competition, just a matter of making the right mistakes! Ultimate proof that it's better to be born lucky than smart.

What do you mean, troll?

[Osvaldo Doederlein]

Java was open sourced under the GPLv2.

http://www.sun.com/software/opensource/java/faq.jsp#g1

I-Opener heaven!

[Anonymous Coward]

Now I can finally run my I-Opener with its native OS!

The I-Opener FAQ is at http://fastolfe.net/2006/iopener/faq if you don't remember the device. It had a very high "cool factor" for its time.

in other words

[josepha48]

QNX is not doing as well as they would like, so they think they can capitalize on open source and maybe take advantage of those who are afraid of the GPL v3.

Yeah that may sound trollish, but there are several companies that are doing the open source thing because they are not doing so well.

I'm not saying this is a bad thing, I'm just saying QNX is not doing as well as I think they would like to.

Re:in other words

[Anonymous Coward]

that being said... the os runs in almost every european car there is...

Re:Microkernel? WTF?!

[evil_aar0n]

Axl's using it to help him write his "Chinese Democracy" album - due out any day now...

Re:Microkernel? WTF?!

[Anonymous Coward]

Let me know when Minix actually supports virtual memory pages. Last I looked it was using segments, which aside from being clumsy, aren't very portable. There are plenty of other more innovative OS's that are much further along than Minix. You could do worse than to look at L4 and Coyotos. You could also do worse than to look at Minix, certainly, but just don't stop there.

Sure, sure

[benhocking]

But the important point is that we could all examine the voting software for bugs.

Sure, sure (correction)

[benhocking]

But the important point is that we could all examine the voting software for undocumented features.

Re:Microkernel? WTF?!

[cerberusss]

Myself, I'm still waiting for GNU/Hurd

Oh right, I heard Duke Nukem forever requires it.

No, the Duke box will actually say "GNU/Hurd or better". That's why I installed Vista.

Screw Off!

[Anonymous Coward]

Get it through your thick heads you fuckwads. We don't CARE about "open source". Only free (as in libre) software is what humanity wants, needs and desires strongly. The fact that you can look at the QNX source but can't really do anything with it makes this a complete non-starter. I don't give a rat's ass about making money. The only thing I care about (and you should too) is making the world a better place with technology, for free, for everyone regardless of race, age, gender, or religion. I want the Muslims to have access to computer technology and software completely free of charge just as much as I think the Christians, Buddhists, Raelians, and the like should. (My only exception is Scientologists because they are not created equal) NO SOFTWARE FOR PROFIT!!! Only FREE software! Who's with me!!!?

Re:Screw Off!

[Anonymous Coward]

I'm with you 100% as soon as food, shelter and the hardware to run your software on becomes free.

Free BSP?

[nurb432]

Cool, that was the one thing that was missing in the past with the free version. All i ever found was ix86. Anything else was $.

They have bills to pay too

[OrangeTide]

Well they do make most of their money off royalties and annual dev kits fees. I think the idea is to get QNX into the academic arena, where you can warp the minds of future engineers and pick up some market share.

Re:They have bills to pay too

[TheRaven64]

You don't have to have your mind warped to see that QNX is a superb design. You do need it warped to want to tie your company's commercial future to a small proprietary software vendor with no second source. It's the business students minds the need to be warping, not the engineers.

Re:They have bills to pay too

[ClosedSource]

"You do need it warped to want to tie your company's commercial future to a small proprietary software vendor with no second source."

What are you saying? IF QNX goes out of business your binaries explode? In an embedded system, the software is the least of your "sourcing" issues. Besides, in the embedded market you shouldn't be tying your company's future to anything beyond your current products. If, for example, you say "We'll be using Linux in our products from now on", you're asking for trouble.

Re:They have bills to pay too

[OrangeTide]

It's not as fast as Linux, and it's not really any better designed than Minix3(BSD licensed now), or Plan9.

It is elegant, no doubt about that. I love writing QNX device drivers :)

Re:They have bills to pay too

[SchwarzeReiter]

Can you elaborate or reference some comparison somewhere? I would be interested...THX.

Minor aside

[jd]

If you hooked a terminate-stay-resident program onto the clock interrupt and used that to page programs in and out, the same way the 4DOS program paged in/out applications on a keystroke, you could have limited multitasking on DOS. However, as noted, nothing was protected, so essentially nothing was safe, but it was possible to use this sort of technique to reduce the limitations of DOS.

Re:Microkernel? WTF?!

[fm6]

I think you've got it backwards. Everyone here is a believer in RMS and RMS is a believer in microkernels. Ergo, microkernels are kewl.

Perhaps you're thinking of the Tanenbaum-Torvalds flamefest, where Tanenbaum argued that the Linux kernel was obsolete from day 1, because it was monolithic. So if you believe that Linus is God (or at least the Flying Spaghetti Monster), then you hate microkernels. On the other hand, monolithic kernels are old technology and microkernels are (relatively) new technology; so no self-respecting geek will have anything to do with monolithic kernels or other paleolithic technology!

It's a very confusing problem. Of course, you can always cop out and say, "I use what works." But what's the fun of that?

Use It for Linux

[Doc Ruby]

Now Linux kernel developers can use code from QNX to make a RTOS that runs Linux apps and the smaller (but still valuable) library of QNX apps. And maybe port Linux (or the new combo) to HW that now runs QNX but not Linux.

It comes too late to finally get Linux working on the Geode-based 3Com "Ergo Audrey" I had such hopes for. But the HW was already obsolete. Maybe now a QNX/Linux OS will get such a promising "home GUI" to work with all the apps that would make it such a neat terminal.

Re:Use It for Linux

[squiggleslash]

Nope, QNX is not going to be Free Software, let alone licensed under anything compatible with the GPL. It will be an infringement of the copyrights of both Linux and QNX to incorporate one's code into the other.

Re:Use It for Linux

[sydneyfong]

I'm not really into the OS business, but I'd suppose QNX and linux were written in entirely different architectures -- one is a microkernel and the other is pretty much based on a monolithic model. Code that works perfectly on a microkernel might not make sense on a monolithic kernel such as Linux, at least without significant changes.

It's not like you can trivially "copy" over some magic code and make an OS more "real time". Just as you couldn't copy over any code that makes an OS more "stable", or "secure". Those attributes are really a consequence of the OS's architecture and design. Of course the kernel developers might gain insight from looking at the QNX source and adopt some of the ideas, but I doubt any significant actual code copying would be useful.

And obviously there is the license issue which makes everything else a moot point. Still, with all the comments saying "oh cool let's copy the code to linux!" it's probably a good thing to point this out.

[Disclaimer again: I know sh!t about computer architecture and OS design]

Re:Use It for Linux

[Doc Ruby]

I don't think kernel development consists of just copying/pasting code between OS'es. And the QNX license won't permit that, anyway.

But Linux kernel developers can now study the QNX code to see the techniques they use to solve realtime problems. And implementing new functions supporting the QNX API will now be more straightforward, in new code. That's what I want to see.

Re:Use It for Linux

[xenocide2]

The techniques are known, as QNX originated from ideas the developers were taught in university. Implementation details are interesting, but several other projects exist with similar goals.

What's more interesting is that with QNX source opened, one can scan for defects via Coverty or other static source analysis tools.

The noncommercial version returns - maybe

[Animats]

The problem is, QNX management has said that before:

"The new QNX initiative consists of several key elements . . .

• In recognition of the ease with which developers can obtain and begin development with Linux, QNX has decided to make their new QNX Realtime Platform free "for non-commercial use". Developers can now download the software and its associated development directly from the QNX website.
• In response to the growing desire for source code that has resulted from the exploding popularity of open-source Linux, QNX will soon release the source code for many QNX applications, drivers, and libraries."

That's from a press release back in 2000.

Last time, the free version of QNX stayed around just long enough that free software developers ported their major packages to QNX. Then QNX management yanked it away.

Despite the new press releases, the QNX CVS source repository hasn't been updated in six years.

Re:The noncommercial version returns - maybe

[NaOCl]

>Despite the new press releases, the QNX CVS source repository hasn't been updated in six years.

The current source publication is not on cvs. It's on an svn server. The web browser access reports the last source update was 14 hours ago.

GPLv3

[OrangeTide]

He is implying that the GPLv3 has restrictions added that make it impractical to use it for commercial projects.

Sweet, now I can...

[rickb928]

...dig that old iOpener out of the basement and, wait, I don't have a basement, damn, I sold the iOpener to some kid, maybe I can score a v5 on eBay, ACK! Only 3, and they obviously don't know which rev they have... dammit, I'm gonna have to work on this, l8r, much, !sleep. Must have.

Re:Microkernel? WTF?!

[TheRaven64]

It's not just unportable outside the x86 family, x86-64 doesn't support segmented addressing, so MINIX 3 will always be a 32-bit OS. For something that's designed as a teaching OS, it has some horrendously unreadable code. Linux is the only kernel I've seen that's worse.

L4 is not a kernel, it's a specification. There are a few nice implementations, but synchronous message passing is just a bad idea. It gives you the worst of both worlds.

I haven't been following Coyotos much since it stopped being EROS. I'm not a huge fan of the capabilities community in general though. Their solution to the problem 'security is too hard for end users to understand well enough to get right' seems to be 'let's make it more complicated.'

If you want an interesting, widely deployed, third generation microkernel, you might want to take a look at Xen. Xen 3.1 is a clean multi-server microkernel design. The fact that the original author's PhD was in lockless data structures really shows in the design.

misleading article, it's not open source

[BokLM]

This article tries to make you believe it will be release under an open source license, but apparently, it will be nothing more than a proprietary license that lets you look at the source code.

It looks like a lame marketing campaign to get people to contribute for free to their proprietary project.

Re:misleading article, it's not open source

[nurb432]

Except you don't have to contribute your work back to them, unlike other 'open' ( but restrictive ) licenses.

Re:misleading article, it's not open source

[BokLM]

Nothing say you have to contribute back your work in BSD or even GPL licenses. You can keep it for yourself. If you distribute it however, the GPL asks you to distribute your changes as GPL, but that not the case for BSD and most other open source licenses. And I doubt that their license will allow you to redistribute your modified version, or that would probably qualify as an open source license (and they say it is not). So their statement that their license is not open source but something better is ridiculous.

Re:Microkernel? WTF?!

[go_epsilon_go]

And I also heard that it's coded in Perl 6

Calling the kettle black?

[ClosedSource]

"And this is why Free Software advocates laugh at people who say 'Open Source' is not ambiguous."

RMS is the king of ambiguity. That's why the phrases "free as in free" and "free as in beer" had to be invented. It should have been called "Freedom Software", but that doesn't have quite the marketing value that "Free Software" does.

Re:Calling the kettle black?

[Hal_Porter]

What about Freedom as in Slavery ;-)

Ok, I'm heading rapidly for Ayn Rand territory by saying it, but it's not completely unfair. Proprietary software lets you get a return on your time investment. GPL software does not - once you release it as GPL other people can provide it for free. So the sustainable cost you can charge for it drops to zero quite fast. Not owning something tends to have that effect. And working for free is slavery.

Re:Calling the kettle black?

[ClosedSource]

You won't get much argument from me. I think all the business plan aspects of GPL'd software are just an afterthought and excuse. Sure, companies like Red Hat can make a profit (at least for now) because 90% of the key work on Linux was done before they invested any money, but if they actually paid for all the early work they'd be out of business.

If and when Linux becomes more mainstream, look for Red Hat's profits to dwindle as the update process gets simpler and people realize that they don't really need a support contract.

Re:Calling the kettle black?

[Daimanta]

It should have been called "Freedom Software", but that doesn't have quite the marketing value that "Free Software" does./quote?

Do you like some freedom fries with that?

Re:Calling the kettle black?

[SiChemist]

So, all the volunteers at habitat for humanity are slaves? And those people that volunteer to clean up the beaches on the coast, they are slaves too? Wow! I didn't know that slavery was prevalent in the US in this day and age. Thanks for clearing that up for me.

I think you might want to crawl out of Ayn Rand territory and join us in the real world where people do things for any number of reasons, not all of them being money.

Re:Calling the kettle black?

[ClosedSource]

Well, we are talking about business uses aren't we? How many people clean up the beach 40 hrs a week for nothing? Besides, if you follow the politically-correct FSF model it's OK to charge for the service of cleaning the beach as long as you're willing to share any modifications you make to the trash with the community.

Re:Calling the kettle black?

[Curtman]

but that doesn't have quite the marketing value that "Free Software" does.

That would be a good thing. "Free Software" still has a stigma associated with "shareware" and other useless crap. Whenever you mention free software to people not familiar with it, it immediately puts you on the defensive about its quality. Most people seem to glaze over when having the free as in freedom discussion. Open source may mean a different thing, but people respond to it much better than free software.

Re:Calling the kettle black?

[dedalus2000]

The money is in customization and the original developers have a edge in the market. there is also a effect on successful projects of amplification of effort. i know plenty of people making money off of gpl'd software.

Re:Calling the kettle black?

[Fyzzler]

All those volunteers along the highway, in the orange jumpsuits voluntarily picking up trash, refute your argument. :)

Re:Calling the kettle black?

[ClosedSource]

I don't believe the evidence supports your conclusion. Look at all the no-cost applications available on the Internet: search engines, email, source control, conferencing, etc. People are happy to use it and don't consider it junk just because it is free. In any case, there's no evidence that a random selection of a GPL'd work is of any higher quality than a random selection of shareware.

Re:Calling the kettle black?

[Curtman]

What evidence? The only evidence I've given you is anecdotal. It has nothing to do with what the reality is. The word "free" to many people equates to cheap crap. There's nothing you or I can do about that except use other words instead. At least to begin the discussion anyway.

Re:Calling the kettle black?

[anothy]

(where's my "misquote>" tag?)

Besides, if you follow the politically-correct FSF model it's OK to charge... ...as long as you're willing to share any modifications you make to the trash with the community.

yup. i think that's an excellent description of what goes on in most Free Software projects.

Re:Calling the kettle black?

[renegadesx]

At least not in China

Complaints Available, NOT RMS sanctioned.

[Anonymous Coward]

"Aye, occasionally an AC does post something worthy after all. I have really had it as well with the "It's not GPL, so it's not really open source" nonsense."*

Sounds similar to what Troll tech did with Qt. And look at the noise that was made back then.

*For a real laugh, observe them when there's an Apache vs IIS discussion. Fair weather friends indeed.

Re:Microkernel? WTF?!

[Anonymous Coward]

Whoa, those jokes are still funny.

Re:Microkernel? WTF?!

[crush]

Nah, round here everyone is a dirty, MacOSX-using, dope-smoking, "web designer" slacker.

Hmmm

[ClosedSource]

"... and overall lack of applications."

It's an embedded OS: It's your job to write the application.

Re:Hmmm

[Anonymous Coward]

I was thinking the same thing. The beauty of QNX is the ease of implementing and testing device drivers. If you have never spent time heavy lifting within Neutrino, you have no idea how sweet the software development life can be. Code it up as a resource manager, and you can migrate your service around your distributed application at will. Ah, I would be so happy to be back designing systems under Neutrino again. Hope to get that chance again soon.

Re:Hmmm

[dargaud]

I know it's been my job for the last 15 years, but still, it's nice when you find, say, a GPL'd FFT library or other useful tools and not have to shell 10 grands for it. If you use Linux as an embedded OS you have more than plenty of choice for tools, libs, etc. No such thing for QNX as closed source draws only commercial apps in its wake.

Re:Hmmm

[ClosedSource]

Well, a library isn't an application. I also don't see the connection between an FFT library and the underlying OS.

Re:Hmmm

[dargaud]

Things go hand in hand. If the OS is free/open, odds are that lots of software under it will be also. If you have to shell 10 grands for the OS (as for some versions of real-time or embedded OSes), you can be sure that the simplest 3rd party library/software will be payware.

Re:Hmmm

[ClosedSource]

Well, you should avoid libraries that are OS dependent if possible. Thus you can use the best combination of tools considering both their cost and their quality. Of course GPL'd code may not be in the running if you wish to keep your own code private.

Not a surprise

[gnalre]

This is not a surprise really. VxWorks the other big embedded OS opened it source a couple of years ago. This was not long after listening to the CEO of Vxworks telling us the vxworks source code was the crown jewels. Well some crown jewels that was.

The truth is all embedded OS have been forced to do this by the rise of linux in the embedded world. Also believe me the difference is huge when you have the source. Wierd behavior and unexplained bugs suddenly become transparent when you can dig into the source. In the end though it doesn't really hurt the vendor since you still pay them for support and development tools.

Download page?

[Anonymous Coward]

I'll believe it when I can click download. So where is the downnload page for QNX ?

Re:Download page?

[TopoGego]

http://www.qnx.com/products/getmomentics/

Re:Download page?

[Anonymous Coward]

That link is not Source code for download. It is more like pre-compiled and need license

Re:Download page?

[NaOCl]

svn checkout instructions, and source browsing instructions are here:

Re:Download page?

[NaOCl]

(2nd try.) The instructions for svn checkout of the qnx os source is here: http://community.qnx.com/sf/wiki/do/viewPage/projects.core_os/wiki/Os_source_guide

Re:Microkernel? WTF?!

[fm6]

Now that is hard to accept, since very few Slashdotters seem to know about the <blockquote> tag.

Re:Excellent news XEN XEN XEN

[Anonymous Coward]

As soon as XEN is ported to QNX, you can have the best of both worlds. QNX realtime OS at dom0 and Linux at domU.

Re:Microkernel? WTF?!

[Brandybuck]

How many successful microkernel based systems are there? One. Only QNX has managed to be successful with a true microkernel architecture.

Now that the source code is open, it would be nice to find out how they got out a robust microkernel years ago while GNU is still struggling with HURD a decade after Stallman belittled Torvalds by saying the kernel was the easiest part of an OS to write.

Cool ! A new Unix system in the open source family

[J4nus_slashdotter]

A lot of systems are based on QNX, such as the new Cisco operating system, IOS-XR.
I'm happy to see a RTOS in the open source family !

Makes verification easier

[Vlaadimir]

When you have to certify a computer to "Design level" as in DO178-B. You need to have the source code for the OS to be able to prove certain characteristics. What this does is remove the negotiation of the cost for the source from the equation. Though this means the company that uses the product will have to certify the OS themselves.

Re:Cool ! A new Unix system in the open source fam

[random44]

I work with a metrology company, and one of our more complicated products uses QNX on the embedded system. A few months ago I was trying to find a workaround for a difficult troubleshooting scenario (that we encounter relatively often). Having access to the source-code should make the job I was trying to do at least possible, if not "easy" (compared to my options before). We pay for our licenses of course, so all this means is we have better options to improve integration!

Repeat the SCO

[Anonymous Coward]

1. "Open" source code of PROPRIETARY_PRODUCT for all to see <----- you are here
2. Start making vague claims that REAL_OPEN_SOURCE_PROJECT includes code of PROPRIETARY_PRODUCT
3. Start program where REAL_OPEN_SOURCE_PROJECT users can license PROPRIETARY_PRODUCT "IP" for $699
4. Select end-users of REAL_OPEN_SOURCE_PROJECT with deep pockets to sue
5. ...
6. Profit!

GPL means "Gay Penguin Lovers"

[newr00tic]

Which means the GPL, (Gay Penguin Lovers,) -license is destructive; it straight out "forbids" quality, by way of its shitty terms, so that anyone with a quality (I.E. non brown) application get scared right off, and seek for saner paths.

Re:GPL means "Gay Penguin Lovers"

[BokLM]

Why do you say that ? Can you explain ?

Does anyone else use QNX?

[mrjodie]

My company has several dozen computers running QNX Windows 4.25 as the embedded OS under our control software. I have to work with it every single day. It's not pretty but it works. I don't know that the OS has ever actually failed. The hard drives and motherboards give out before the software. I personally think opensource on this is more of a win for the consumers, because the developers just don't care, any more.

I will also admit that all of our new systems run Windows XP Embedded. Yeah, I designed and built a better system. That was me.

Re:Microkernel? WTF?!

[crush]

The OK Labs approach is interesting too and they specifically claim that their approach is secure. The latter paper has good historical background on the history of microkernels (which you may already be aware of) and mentions some that I hadn't heard of including IntegrityOS which is supposedly widely deployed in the military. Worth a read.

Re:Microkernel? WTF?!

[Anonymous Coward]

the kernel was the easiest part of an OS to write He was correct, if you are talking about a monolithic kernel.

The reason everyone and their brother doesn't have a viable microkernel, is that microkernels are hard to write (and get good performance). Writing a poorly performing microkernel is easy, and that's why they have the reputation of being slow.

There is something unstated here though; while writing a monolithic kernel is easy, keeping it running (maintaining it) is hard (because of the lack of protection).

              writing maintaining

mono easy hard
micro hard easy

The reason most computer scientists prefer microkernels is that they are taught to prefer s/w architectures that have the lowest cost over time, and microkernels clearly win in the long term. Unfortunately, as everyone knows, outside of academia, decisions are rarely made with the long-term in mind.