Shared Source – More than Windows.


Thanks for the feedback on the first posting with meat on the bones. I’m moderating the comments in an attempt to keep the spammers at bay – but I’m posting all comments, positive or negative to keep this a space of “open” thought. Sorry, couldn’t help myself.

 

I’m going to repeat myself for the moment; given some of the comments, this bears repetition.

 

Shared Source includes more than the Windows programs. Within a business context, Windows is arguably the most valuable intellectual property in all of software. We are sharing that source code with a great deal of care. Our reference license is not meant to be analogous to any OSI license. (The big hint was when we called it “Shared Source” instead of “open source.”) The dispassionate view of understanding how platform source code is used and how our customers and partners want to work with it was how we arrived at this approach. Thus, the fact that well over 99% of those developing on Linux will never wish to modify the source becomes an important data point. I give full credit to those who do – tackling a source base representing millions of lines of source is a major undertaking no matter what OS or app you are talking about.

 

Shared Source includes more than the Windows programs. The challenge for any organization (and Microsoft is no different) is in understanding what the collaborative development model can do for that specific organization. There is no one-size-fits-all approach.

 

Shared Source includes more than the Windows programs. Looking across the spectrum of technologies that we produce, there is an enormous opportunity to reach out more effectively to the developer community. We have engaged communities well for years; witness MSDN, TechNet, the Visual Basic dev community, etc. But we always have room to learn more and to improve. We are learning from open source and thinking deeply about how that model can either blend or, at times, conflict with what we are doing.

 

Shared Source includes more than the Windows programs. I have categorized our approach into four community-centric concepts:

  1. Supporting existing customers
  2. Encouraging new development
  3. Enabling academia and research
  4. Creating opportunities for our partners

Any source sharing we do falls into one of these broad categories. I dismiss the idea that, if you are not talking about a core asset or a platform technology, then there is no benefit from source code access.

 

Shared source includes more than the Windows programs. Developer tools, samples, snippets, app-layer technologies – all are fair game. I’m not interested in source licensing as a religious battle, nor am I interested in hamstringing my organization by limiting our approach to a single license or technology. Shared Source is predicated on a spectrum approach – licenses, technologies, and communities. The open source community would be really boring if there were only one project – an operating system.

 

This posting is provided "AS IS" with no warranties, and confers no rights.

Comments (5)
  1. Lon Ingram says:

    First, as an Airman in the US Air Force, I sympathize with these Microsoft punks. It is frustrating to be a member of a subset of a larger organization that is often excoriated as a whole for the failures and mistakes of a few.

    That being said, I think that this post is a bit disingenuous and disappointingly so. I’m not really sure that anyone is really asserting that "if you are not talking about a core asset or a platform technology, then there is no benefit from source code access". I think that people are saying that core asset and platform technology source code access is more important and vital than access to Halo source code because of its very "core" and "platform" nature.

    I don’t want access to Windows code so that I can release LonDows 2010, I want it so that better programmers than I can help MS make a better product. The point is that ALL other source code depends on the OS code and we would all feel better if we knew that anyone can take a look at it without fear of NDA lawsuits and such. Reading your last two posts tells me that argument by false analogy is beneath you, Mr. Matusow. Don’t tell us that ya’ll won’t release WinXP code because it isn’t important. Don’t tell us that ya’ll won’t release it because your XML engine code is just as important. It isn’t and you know it.

    Ya’ll won’t release it because you are afraid of the financial consequences. Admit that and the credibility of Shared Source may be a bit more. That credibility is an uphill battle, but one I sympathize with. Nobody believes the military when we just flat mess up and no one will believe you either. Way it goes.

    On the other hand, I celebrate any embrace of OSS by MS. Ya’ll seem to have finally figured out how to make a world-class OS with XP (historical contributions notwithstanding), so don’t be afraid to strut your stuff.

    On a more serious note, I think it is swiftly becoming an irresistible responsibility to reveal your OS code to all. Windows is now the standard for the world. That is not going to change any time soon no matter how many times I install Debian on my Thinkpad 600. More importantly, that won’t change anytime soon if I could install DebianWinXP on my Thinkpad.

    So. What do I want? I want MS to release their OS code for real. Not to select groups, to everyone. Ya’ll will still be best positioned to sell updates and expansions. Just let us in on the core technology. Yes, a </irony>small</irony> request, I know. I do think that ya’ll will be the better for it, though.

    From the Shared Source site:

    "Microsoft source code components” – what the h#ll does "source code components" mean?

    ///DISCLAIMER///

    In NO WAY should any comments in this post be taken, interpreted, construed, or otherwise accepted as those of the United States Air Force, the Department of Defense, or the government of the United States. All statements and opinions are mine and mine alone. No lawsuits, monkeyboys.

  2. Andy O says:

    This is going to be "amusing". Are we to witness you do post after post after post in which you try to argue that it smells and tastes like openness or freedom, so therefore it must be?

    What you fail to see (or deliberately ignore) is, that so much of what your organisation delivers, is so incompatible with what fans of openness and freedom likes, that it cannot be compared.

    You will only be like a jester walking through the room banging metal and honking horns, in the eyes of those who already knows the value of real openness, collaboration and common pools of knowledge.

    I suspect you will succed in at least muddy the waters for outside spectators, that not yet have enjoyed the freedoms you are trying to freeride and wrap in your own colors..

    Your example with the ratio between users and code-tweekers of open source is a classic example of misguidedness. It isn’t about the ordinary user being able to hack the kernel. It is about trust. It is not that I can view and change the code, it’s that anyone can. I belive that there is so much stuff in windows that wouldn’t survive for seconds, if the users could remove it. As long you don’t see the connection, your blogging about the subject doesn’t really matter.

    And about the "free" part. Here is a burp a nice person wrote:

    "People are sick of license terms that treat them like criminals, where even when you try hard to obey, you never feel free of that worry…am I allowed to do this? They love GNU/Linux because you can share with your friends and family freely, install it on as many computers as you own at home and at work. Sick of saving proof of purchase certificates under pain of a visit from the IP police and fines when they can’t find that piece of paper from 1998. Sick of typing in numbers to prove they bought the software, and having software call home to validate their right to use what they bought, and companies that shove one-sided EULAs down their throats, claiming the right to monitor their hard drive for compliance. Sick of businesses that care about money for themselves first and customers a distant second."

    You might not get it and you might not believe it. But before you turn away from building taller and taller walls around the users… (All of them, not only those that has signed away their soul to gaze at the windows-scriptures) this blog is so insignificant to the core of the discussion that it borders on the amusing.

  3. In a barely-related question: Is there somewhere a matrix available of what MS source-code is available, and under what license, to whom? For example, if I am interested in writing an XP video/display driver under the GPL (or some substantially similar license) (which I am), what source code can I look at without tainting myself, assuming I am not willing to spend outragous amounts of money, for someone who is not an academic nor corporate?

    What if (and this one is hypothetical), I’m a hospital, which wishes to verify a random subset of the code used in it’s systems, and has money to spend on it? What code is available to me?

    I count 18 seperate shared-source programs on http://www.microsoft.com/resources/sharedsource/Licensing/default.mspx, most of which aren’t clear in the short descriptions who they’re aimed at, or what source is available in which of them, under what sort of license.

  4. tecosystems says:

    I finally got a chance this morning to catch up with Microsoft’s Jason Matusow, better known as Redmond’s point person on all things open source, and one of the topics of discussion was the perception that Shared Source is just…

  5. The Shared Source License distributed with IronPython 0.7 includes multiple references to the phrase "that read on." Though this phrase does appear in approx 130 patent-related found via Google*, I do not understand what it means. Can anyone help?

    *

    http://www.google.com/search?hl=en&q=+%22patents+that+read+on%22+OR+%22patent+claims+that+read+on%22&btnG=Google+Search

Comments are closed.

Skip to main content