The New Linux: OpenStack aims for the heavens In late November, Mark Collier brought his OpenStack crusade to the Far East. Collier oversees community relations for the OpenStack project, an ambitious effort to create a completely open source ecosystem for building Amazon-like "infrastructure clouds". After the Thanksgiving holidays, he and his team flew to Beijing and then …
I see that MS is said to be enthusiastic about this project and that rang alarm bells. Now it is well known that MS is deeply averse to the GPL so the question is what license has this project been released under?
The answer is Apache 2 see:
http://www.theregister.co.uk/2010/07/19/nasa_rackspace_openstack/.
Now the reason NASA ditched Eucalyptus and started NOVA was because of the "open core" aspect of Eucalyptus meant that they were unable to contribute code to increase the scalability of Eucalyptus. By adopting an Apache license thay have gone to the other extreme and basically given away all their work to any proprietary shark that wants to freeload on the tax payers' dollar and take the whole shebang private. No wonder MS is keen.
Another wasted opportunity to make the project truely free and open. Bah!
"By adopting an Apache license thay have gone to the other extreme and basically given away all their work to any proprietary shark..."
I like how you tried to make a free source license look like an evil conspiracy by adding the phrase "to any proprietary shark". We're talking software, not comic-book secret science.
If Proprietary Shark, Inc. adds a feature that isn't in the original source, are you really saying that it becomes impossible to duplicate in open source code? Isn't that what open source software does?
Hmm, and which license did MySQL use? Why, it looks like GPL (<http://news.cnet.com/2100-7344_3-5173014.html>)!
Hm, that indirectly reminded me of another point. I might be misunderstanding the situation, or this might fall in some funny legal loophole, but... How *could* NASA put a license on *anything* to begin with!? I thought everything they did was supposed to become automatically part of public domain, by US law. Am I wrong? I probably am, since not a lawyer and etc.
I take that from the "license" header from NCBI's software, which reads in part:
"this software is a "United States Government Work" under the terms of the United States Copyright Act. It was written as part of the authors' official duties as United States Government employees and thus cannot be copyrighted. This software is freely available to the public for use. The National Library of Medicine and the U.S. Government have not placed any restriction on its use or reproduction."
One of NASA's pages seems to suggest something on those lines, but it's ambiguous: http://www.jsc.nasa.gov/policies.html
They could perhaps accept contributions from other people (non-employees) under the licence of their choice. If the outside contributions are mixed in with the employee-written code then the whole lot is effectively under the licence they chose.
Also, "United States Government Work" may be public-domain in the USA, but what about elsewhere?
Some of us who like open source actually prefer Apache and BSD licenses, irrespective of whether we want to develop closed source derivatives or not.
If I were to release code I wrote (mostly a question of having clean code of general interest), I would most likely not choose GPL.
Like you, I assume M$ sees closed code potential, but better an onboard M$ which hoards its crud than a project with limited vendor buy-in (think "codec" if you want to see what I mean).
If the project takes off, M$ may yet be taken to task for then introducing lock-in to its users in its implementation. And savvy users will stay away from the M$ flavors.
MSFT's TCP stack is almost entirely open source ( BSD) - would it have been better for them to have 'done a DEC' and implemented their own proprietry version of the internet?
If using open source components without them giving anything back means that MSFT at least stays a little more compatible with open standards that's fine
By this do you mean DECnet? This was originally developed in the late 1970's. exactly how much TCP/IP was there in commercial use then? Not a lot.
SNA ruled the roost back then. DECnet was a very popular alternative even for IBM shops. I remember using DECnet to link to stonking great BT Mainframes together (one IBM and one Amdhal) as it was a proper peer networking unlike the SNA hierarchial model.
DECnet Ph IV followed the OSI 7 layer stack model. That is hardly a propretary version of the internet is it?
In the end, you could use DECnet over TCP/IP. The LAT protocol was far more efficient for character transmission than TCP. Running DECnet over X.25 was very simple.
At least with DECnet, when you did a read and it completed, you knew you had everything.
" By adopting an Apache license thay have gone to the other extreme and basically given away all their work to any proprietary shark that wants to freeload on the tax payers' dollar and take the whole shebang private. "
What kind of "argument" is that? Even if MS were to "embrace and extend" OpenStack, it won't stop others from using the official OpenStack data and competing with their *own* contributions to the project.
I keep reading variations on this bizarre logic from pro-GPL advocates, but they've never once managed to provide any evidence that this anything other than a straw-man 'problem'.
Even Public Domain greatly predates the GPL, but it's still working just fine today. Personally, the only reason I can see for such vocal advocacy of GPL licensing is egotism. It's like giving away a million dollars to charity, while running a million-dollar PR campaign to make damned sure everyone knows about your "utterly selfless" gesture.
--------
' " By adopting an Apache license thay have gone to the other extreme and basically given away all their work to any proprietary shark that wants to freeload on the tax payers' dollar and take the whole shebang private. "
What kind of "argument" is that? Even if MS were to "embrace and extend" OpenStack, it won't stop others from using the official OpenStack data and competing with their *own* contributions to the project.'
--------
I expect the worry is not that extensions will be added that can compete on merit - but that some entity will introduce them under highly restrictive conditions (if at all) and use a dominant position in a particular market to make them a de facto standard; a 'standard' which others cannot comply with. The sentence quoted does not specifically mention Microsoft and, although they could arguably make an easy "go to" villian in the first instance, the argument is applicable to anyone with sufficient leverage.
--------
'I keep reading variations on this bizarre logic from pro-GPL advocates, but they've never once managed to provide any evidence that this anything other than a straw-man 'problem'.'
--------
I don't think this is just limited to pro-GPL advocates (or supporters of any other licensing style) and I believe there is plenty of evidence (both direct and circumstantial) of this happening; with respect to Microsoft you may care to remember e.g. some of the details of the DoJ vs Microsoft anti-trust trial or the Halloween documents (c.f. the discussion of the "de-commoditization" of protocols and applications).
--------
'Even Public Domain greatly predates the GPL, but it's still working just fine today.'
--------
I'm not sure this sentence really makes any sense - perhaps you could eloborate ?
--------
' Personally, the only reason I can see for such vocal advocacy of GPL licensing is egotism.'
--------
I agree that the noise coming from people that seem believe that the GPL is the only possible license for any situation is irritating but, alas, the most vocal supporters of a belief are often not those most qualified to explain it - and I don't believe that is limited to any particular 'side' of any argument. That said, i'm sure that few people - either for or against the licensing style used in the GPL - believe that egotism is really is the only reason for advocating it.
while largely and unfairly ignoring FreeBSD, OpenBSD, NetBSD and other like that. They are all powerful, robust, highly secure but they are rarely mentioned. Although their license is even more liberal than GPL and cost the same, Oracle, IBM, Intel, Google and other big names in the computing industry considered Linux instead. More than that, in Microsoft's opinion GPL is a cancer, is un-American, is communist, hell they can't even spell those three letters without giving them hive while BSD, Apache and other licenses are OK.
Yes, pro-GPL advocates are vocal because they do mind whenever someone tries to steal their work in the same way that Microsoft would be angry if you don't play nice with their license.
The only thing that really upsets the stomach of all those anti-GPL proponents is that this license ensures a perpetually level playing field. Maybe this is not important for you but for some of us it is.
If I've given my work for free, everyone else must too!
Yes, if I code for free it is my personal choice. BUT EVERYONE MUST MAKE THE SAME PERSONAL CHOICE. I DEMAND IT!!!!
</sarcasm>
What the GPL junkies don't get is that it IS personal choice, and some people have different priorities.
The GPL is good because it allows coders to demand "quid pro quo", so they give and they get. Personal choice.
Even freer licenses are also good because the allow coders to say that they don't need quid pro quo -- if they are confident that they're getting enough out of the code for their own benefit, they don't need any quid pro quo.
NASA is acting in public benefit, so they don't care about "sharks" -- they're doing what they've always done with their research (cold war secrecy excepted) and making it available to others. It's a public research group, and that's how these things work.
Rackspace is a storage/hosting provider, and to them, the main goal of the software is to get people to use more storage and bandwidth. Widespread availability of the platform facilitates this.
So, yes, both parties CHOSE to go for a "very free" license, because it is in their interests.
I find it interesting that random internet commentards think they know Rackspace and NASA's business better than Rackspace and NASA. How many degrees do these commentards have? Less than the minimum CV for employment at NASA, I'd guess.
The problem with using an Apache license as opposed to GPL is that a company like Microsoft could for instance sell "Windows Cloud Stack", using all the Open Stack code plus some closed code of their own to support Azure or other of their products.
While preventing anyone in the Open Stack comunity of developping an open alternative to the close code by the use of patents.
Where you read Microsoft you can also read VMWare, Oracle, Amazon... If I was an american tax payer I'd be "mighty pleased"
You''re all very confused about why microsoft supports this project. Their end of the deal is to support their hypervisor - this means the more people use their part the more they sell of their own OS, if you're microsoft this is no bad thing. Given 'enabling clouds' (wow do I hate that term) isn't their core business they'll always be happy to open source their code. It will enable other people to do the same thing with their cloud platforms with a nice handy open source reference/example which again for microsoft is no bad thing.
Ignoring the fact Microsoft isn't new to open source for a moment - the motive is obvious and shouldn't 'ring alarm bells'.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
