Surprises in the Windows 7 license agreement

Surprises in the Windows 7 license agreement

Summary: Over the weekend, I had a chance to spend some quality time with the leaked Windows 7 build that’s zipping around the Internet. No one at Microsoft will confirm whether this code is the same as the beta due to be officially released in early January, but it bears every earmark of being the real thing. One of the first things I did before installing the software was to read the end user license agreement (EULA), carefully. Most of it was boilerplate, but I found a few surprises hidden within the legalese, including an expiration date that suggests a summertime release.

SHARE:

Over the weekend, I had a chance to spend some quality time with the leaked Windows 7 build that’s zipping around the Internet. No one at Microsoft will confirm whether this code is the same as the beta due to be officially released in early January, but it bears every earmark of being the real thing.

One of the first things I did before installing the software was to read the end user license agreement (EULA), carefully. Most of it was boilerplate, but I found a few surprises hidden within the legalese, including a revision code at the end, “EULAID:Win7_B.1_PRO_NRL_en-US,” which indicates that this is indeed Beta 1. Here’s a summary of some other interesting additions:

  • You can install as many copies as you want. The agreement specifically waives any restriction on the number of copies you can install:

    You may install and use any number of copies of the software on your premises to design, develop and test your programs for use with the software.

    I expect this wording is from a build specifically released to software developers. This wording might change to a more general “for evaluation only” clause in the public beta release.

  • Don’t use it in a production environment. That’s generally good advice for any product with a beta label on it, but in this case it’s explicitly covered in the agreement:

    You may not use the software in a live operating environment unless Microsoft permits you to do so under another agreement.

  • The software expires on August 1, 2009. Although I’ve read reports from other testers of a different expiration date, the copy I looked at includes a “Time-Sensitive Software” clause that reads in part: “The software will stop running on August 1, 2009. You may not receive any other notice.” That timeout date adds further credence to the notion that the final release will be ready in May or June.
  • It’s OK to install in a virtual machine. The license agreement for the original release of Windows Vista includes some truly opaque wording about installing in a virtualized environment. This wording was significantly cleaned up for the Vista SP1 license agreement, and this same language appears in the Windows 7 EULA. The “Use with Virtualization Technologies” section is straightforward:

Instead of using the software directly on the licensed device, you may install and use the software within only one virtual (or otherwise emulated) hardware system on the licensed device.

  • You’ll need to take a few extra steps to lock down your privacy. In section 4, the license agreement specifically notes that some features that normally require you to opt in are instead turned on by default:

    Because this software is a pre-release version, we have turned on some internet-based features by default to obtain feedback about them […] You may switch off these features or not use them.

    Most of the services on the list are fairly benign and involve little risk of divulging personally identifiable information. However, if you work with sensitive data files you might want to turn off the Customer Experience Improvement Program and automatic error reporting options.

  • Activation and validation are alive and well. Anyone who was hoping that Microsoft would back off from its hard-line antipiracy initiatives might be disappointed. The license agreement specifically describes activation and links to a privacy statement that says activation is required for Windows 7. The lengthy section on validation is identical to the one in Vista SP, including the bold-faced warning:  “You are not permitted to circumvent validation.”
  • No benchmarks allowed. As in previous beta releases of Microsoft operating systems, the license agreement includes a prohibition on speed tests:

    You may not disclose the results of any benchmark tests of the software to any third party without Microsoft’s prior written approval.

Of course, most people don’t bother to even read license agreements, so it’s unlikely that this one will be followed strictly. In fact, I suspect that some people have been clicking their stopwatches feverishly over the past few days in anticipation of another round of benchmark results, regardless of what the EULA says.

Topics: Windows, CXO, Legal, Microsoft, Software, IT Employment

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

85 comments
Log in or register to join the discussion
  • Benchmarks would be pretty silly at this point.

    You can expect that a lot of the core OS is dumping detailed logs in circular files, or armed to produce detailed core files that hinder performance, etc.

    Do you think the "you can install on as many machines as you want" will make the production EULA? :D

    TripleII
    TripleII-21189418044173169409978279405827
    • Come on ...

      [i]Do you think the "you can install on as many machines as you want" will make the production EULA?[/i]

      Now would that be as many machines "per user" or "per seat" or "per owner" or "per organization" or "per vendor" or "per $INSERT"?

      I get really confused with these worth as much as the paper they are printed on EULAs. ;)
      MisterMiester
    • Benchmarks show

      (Obviously violating the EULA) Benchmarks show that Win7 B1 is faster than either earlier builds or Vista SP1 (which is as fast as XP SP3). If the Beta is slowed down, then Win7 will be really fast.
      daengbo
      • You ar right, but can't draw final conclusions.

        For all we know, major subsystems are simply stubbed and to be added, etc, so it works both ways (or could). Either way, until the RTM build is released, fuzzy benchmarks are OK to show trends, etc, but you can't rely on them.

        TripleII
        TripleII-21189418044173169409978279405827
      • RE: Surprises in the Windows 7 license agreement

        The most superb clue for a <a rel="follow" href="http://www.pikaba.com/l/cars-trucks/used-cars-acura/seattle.aspx">Used Acura Seattle</a> trouble is certainly <a rel="follow" href="http://www.pikaba.com/l/cars-trucks/used-cars-honda/seattle.aspx">Used Honda Seattle</a> and <a rel="follow" href="http://www.pikaba.com/l/cars-trucks/used-cars-toyota/seattle.aspx">Used Toyota Seattle</a> repair.
        MR.Picaba
  • Why? I really wonder...

    "[B]# No benchmarks allowed.[/B][I] As in previous beta releases of Microsoft operating systems, the license agreement includes a prohibition on speed tests:[/I]

    [U]You may not disclose the results of any benchmark tests of the software to any third party without Microsoft?s prior written approval.[/U]"

    Seems odd to me... ]:)
    Linux User 147560
    • Standard stuff

      The software is developed for features, not speed, at this point. So as TripleII points out, the benchmarks would be pretty silly.

      Nonetheless, people do publish benchmarks. I would be much more impressed with someone who actually consulted with Microsoft during the preparation of any tests to find out whether there are specific issues that might affect test results. But so far, I have not seen anyone perform that basic step, which used to be commonplace years ago.

      I'm not talking about getting permission, only with getting information.
      Ed Bott
      • Benchmark Kiddies

        You are right Ed. These days "benchmarkers" seems to
        be clueless journalists with agendas (looking at you,
        Kennedy) who only know how to double-click an
        automated benchmark suite.

        The most common mistake I have seen is the failure to
        realize that both Vista and (I assume W7) will try to
        use the GPU to achieve the best balance between FPS
        and graphical quality. This means that Windows may
        switch on features of the graphics adapter which will
        lower FPS. Generally FPS above 70 is a waste of effort
        and if FPS can be traded in for improved graphical
        quality it makes a lot of sense.

        Vista supports DX10, Windows 7 supports DX11. In none
        of the benchmarks I have seen have the "researcher"
        even considered this. If W7 switches on DX11 features
        (improves quality) and that causes FPS to go down
        because, it becomes meaningless to compare FPS in
        "game benchmarks". Unless the researcher can guarantee
        that the game was executed with the exact same
        rendering settings in both instances.
        honeymonster
        • *sigh*

          That's not how DirectX works. Applications requests a specific version when you instantiate it (Direct3D10Create...). There are no benchmark programs that support DirectX11, so they're all going to be using the DirectX10 API.

          The OS *cannot* override this and will not add DX11 features to a DX10 program. The OS will *not* turn on new functions at the expense of framerate as that will TRASH backwards compatiblility more than it will help. Only optimizations in the pre-existing APIs functions would have any effect, which could only *help* the W7 benchmarks.

          Also, the OS does not override the other settings (resolution, aliasing, etc) requested by the software. You can override them at the driver, but that's not the default.
          3vi1
          • condescending jerk

            you gave a good technical answer
            but whenever someone responds with 'sigh' they're being a condescending arrogant jerk

            so -1 for you
            thissitesux
          • I disagree...

            I for one fully understand the sigh; when someone is being pedantic, it's one thing. But the French have a word, "cuistre", which more or less means "a pedantic orator who doesn't understand his oratory topic of choice and embarrasses his audience through his lack of knowledge", and IMHO someone like that certainly deserves a sigh.

            +1 from me
            914four
    • No surprise you would find something common odd.

      This is beta software and you find it odd? It's also not that odd even for released software.
      ye
      • I can't imagine that clause would hold up in court

        I agree that benchmarks in a beta are not useful data, but even though that clause does appear in production releases as well, it has about as much legal standing as the EULA itself. Perhaps even less, because this is taking away an actual constitutional right (freedom of speech) without signed consent from the consumer. Most of the other restrictions of a EULA do not take away constitutional rights as this one does.
        Michael Kelly
        • Aimed at competitors

          I agree that trying to enforce this against the press would be pointless and morally wrong.

          Historically, the point of this sort of license agreement is to stop competitors from making claims about the product. This sort of clause used to be common in database products and may still be. The idea was to prevent the competition from claiming their product is faster than yours.

          My guess is that this license provision would be much more defensible if the action were against a competitor.
          Ed Bott
          • I don't agree with that either

            I don't see anything wrong with a competitor making factual statements about their products. Now if they were to make non-factual statements or were to draw non-factual conclusions based off misleading benchmarks then that would be a different story. But that would also be liable and/or slander, and you don't need a EULA for protection against that.
            Michael Kelly
          • Commercial speech

            Commercial speech is subject to many more restrictions than ordinary speech.
            Ed Bott
          • lawyer fulltime employment act

            The problem with benchmark claims is that while the actual facts may be true the conclusion might be false. The only thing that can be done is no matter what the claim is, you immediately sue to get the claim quashed at least until you can vet it. The difference is with a eula the issue is clearcut, without it, you have to actually go thru a trial.
            _JimB_
          • How can having it in the EULA make it clear cut

            when the EULA itself is not clear cut? The EULA is not a signed document nor a verbal agreement nor a legal statute. I do not see any federal judge revoking anyone's freedom of speech without any of those.
            Michael Kelly
          • Factual but misleading is the issue.

            It prohibits a competitor using benchmark results as they could claim

            "Benchmarks show Vista slower then our X-X-X-X"

            That is the factual statement they are required to make, and may be 100 percent accurate, without the information that it was a Beta vs RTM version being benchmarked.

            With that info left off the consumers is mislead to believe both sets of numbers are from the versions that sits on the store shelves that they will be purchasing.

            And in the end the actuall finished Vista RTM vs X-X-X-X RTM benchmark could show that Vista is actually the faster of the two
            GuidingLight
          • A beta is not an actual product

            Claiming a benchmark on a beta rather than the released product is like an auto making claims about a test model before the actual release of the car. So no it is not factual data to release the beta benchmark rather than the actual product's benchmark when comparing to another actual product.
            Michael Kelly