Court Orders Microsoft To Stop Selling Word



  • The story goes:

    Back in August, the Canadian company i4i sued Microsoft over a patent infringement on some sort of XML processing feature in MS Word. Microsoft lost that case and was ordered to pay i4i $290 million and to stop selling versions of Word which infringed on the copyright. Microsoft appealed, but the ruling was upheld. You can find i4i's patent here: http://i.zdnet.com/blogs/msfti4icomplaint.pdf (it's more programmer jargon than legalese, so it isn't too difficult to read).

    Basically, the patent is for an algorithm which builds a table of
    the character positions of the tags within an XML document relative to
    the content. The XML tags themselves aren't counted when counting the
    character positions. This makes it easier to split the tags and content
    into two separate files, or to combine these tag and content files back
    into a single XML document.

    That's it. That's the whole patent, summed up into a single paragraph - and it's worth $290 million! The patent isn't specific to XML either - it applies to any structured content document that mixes tags/metadata with data.

    I just can't believe Microsoft lost this case. Judges have no clue when it comes to technology (and I'm surprised Microsoft's lawyers didn't succeed). 

    EDIT:

    Why is the "Insert Hyperlink" button disabled? I have to go into HTML view and manually add the <a> tags to this post.



  • I nearly made a mistake and clicked "REPORT ABUSE", thinking it meant report abuse of the patent system. 



  •  You're about four months too late. We already had this discussion.



  • @amischiefr said:

     You're about four months too late. We already had this discussion.

     

    Four months ago, Microsoft was trying to appeal the ruling. I posted this mainly to say that they lost again, which I find surprising. Damn, I should have made that clearer. 

    Plus, this is the first time I actually bothered to carefully read the patent for myself. I was mistaken four months ago - the patent isn't on a file format (don't know where I read that it was), it's on an algorithm for processing and splitting documents. Now that I understand it, it just seems worse. 



  • @amischiefr said:

    You're about four months too late. We already had this discussion.

    And the funny thing is, he's posting the same bullshit again.  I don't know the technical merits of the patent, but several sources I trust have said that i4i isn't a patent troll, that they have a legitimate claim on the technology they spent much time and money inventing and that Microsoft ripped off.

     

    Seriously, Huf, lay off Slashdot for awhile.  Your paranoid fantasies are becoming really pathetic.



  •  Patent troll, no.  Bad patent, yes.  Ideas shouldn't be patentable.  Implementations of ideas should be.

    The defense of a patent should never be "They're doing something similar to what we're doing."

    Even from the point of view of what was considered state-of-the-art when the patent was filed, it appears to be a logical (obvious) extension of then-current standards.

    As such, it would seem that XML as a whole (or at least many implementations thereof) would infringe on i4i's patent.  The patent is for the definition and storage of metatags.  So basically, if you are using XML to store data, you cannot allow the end-user to define custom fields within the XML without paying royalties to i4i.

    Of course, i4i could also find themselves in a sticky situation by suing the company that makes the software that their software depends on  (nearly all of their products are add-ons for Office), and Microsoft is under no obligation to license software to i4i (businesses have the right to refuse service except in certain civil rights circumstances, and I don't believe that being hostile to the business is a protected exemption).



  • @smbarbour said:

    Patent troll, no.  Bad patent, yes.  Ideas shouldn't be patentable.  Implementations of ideas should be.
    Ideas are not patentable. Only implementations of ideas (such as devices and methods) are.

    The defense of a patent should never be "They're doing something similar to what we're doing."
    And it wasn't in this case. It was "they're doing exactly what we patented", specifically separateing a structured document's structural tags and content to facilitate editing them independently. Note that the patent didn't cover any method of editing the metadata separate from the data, just the more efficient means of doing so by storing them inpendently.

    Even from the point of view of what was considered state-of-the-art when the patent was filed, it appears to be a logical (obvious) extension of then-current standards.
    Sadly, Microsoft didn't start making those kinds of arguments until too late in the legal process.

    As such, it would seem that XML as a whole (or at least many implementations thereof) would infringe on i4i's patent.  The patent is for the definition and storage of metatags.  So basically, if you are using XML to store data, you cannot allow the end-user to define custom fields within the XML without paying royalties to i4i.
    Nonsense. It is specifically parsing a structured document such that the structure and the contents are stored independently. You can still store the metatags along with the data. You can still allow the end-user to define custom fields without paying royalties.

    My generic response to this kind of argument is this simple:

    If they patented something useful and important, just wait until their patent expires. Then it's yours to do whatever you want with for free. Be glad patent law required them to disclose it.

    If they patented something useless and unimportant, just do something else until the patent expires. It shouldn't be much of an inconvenience.

    This falls under the second part. Just do something slightly different until the patent expires. No big deal.



  • @joelkatz said:

    Ideas are not patentable. Only implementations of ideas (such as devices and methods) are.

    What about algorithms? Algorithms are ideas for how to solve a problem, whereas an imlpementation is a piece of code. What i4i recieved a pantent for was an idea.

    My generic response to this kind of argument is this simple:

    If they patented something useful and important, just wait until their patent expires. Then it's yours to do whatever you want with for free. Be glad patent law required them to disclose it.

     

    A patent lasts 20 years and this one was issued in 1998. I have to wait 9 more years before I'm legally allowed to record the location of tags within an XML document, so I can strip them out and add them back in as I want (unless I invent a weird, convoluted way of doing so - because i4i patented the obvious way). That's unreasonable.



  •  Does this count in Europe too? I don't know what kind of software i4i makes, but I'd certainly like to buy a Word 2kx copy.



  • @Huf Lungdung said:

    @joelkatz said:

    Ideas are not patentable. Only implementations of ideas (such as devices and methods) are.

    What about algorithms? Algorithms are ideas for how to solve a problem, whereas an imlpementation is a piece of code. What i4i recieved a pantent for was an idea.

    The term 'algorithm' covers a wide range of possible things, from abstract ideas to very precise recipes. Not all agorithms are patentable.

    My generic response to this kind of argument is this simple:

    If they patented something useful and important, just wait until their patent expires. Then it's yours to do whatever you want with for free. Be glad patent law required them to disclose it.

     

    A patent lasts 20 years and this one was issued in 1998. I have to wait 9 more years before I'm legally allowed to record the location of tags within an XML document, so I can strip them out and add them back in as I want (unless I invent a weird, convoluted way of doing so - because i4i patented the obvious way). That's unreasonable.
    It is 20 years from the data of application, not the date of issue. In any event, the fact that you find this one massive outlier of a patent slightly inconvenient is really meaningless. Yes, there will be the occasional patent that will minimally inconvenience you. (And, in fact, this patent is likely invalid anyway. Microsoft just b0rked the procedure, failing to challenge it based on the right prior art.)



  •  @joelkatz said:

    @Huf Lungdung said:

    @joelkatz said:

    Ideas are not patentable. Only implementations of ideas (such as devices and methods) are.

    What about algorithms? Algorithms are ideas for how to solve a problem, whereas an imlpementation is a piece of code. What i4i recieved a pantent for was an idea.

    The term 'algorithm' covers a wide range of possible things, from abstract ideas to very precise recipes. Not all agorithms are patentable.

    My generic response to this kind of argument is this simple:

    If they patented something useful and important, just wait until their patent expires. Then it's yours to do whatever you want with for free. Be glad patent law required them to disclose it.

     

    A patent lasts 20 years and this one was issued in 1998. I have to wait 9 more years before I'm legally allowed to record the location of tags within an XML document, so I can strip them out and add them back in as I want (unless I invent a weird, convoluted way of doing so - because i4i patented the obvious way). That's unreasonable.

    It is 20 years from the data of application, not the date of issue. In any event, the fact that you find this one massive outlier of a patent slightly inconvenient is really meaningless. Yes, there will be the occasional patent that will minimally inconvenience you. (And, in fact, this patent is likely invalid anyway. Microsoft just b0rked the procedure, failing to challenge it based on the right prior art.)
    Patents resulting from applications that were pending on June 8, 1995
    can last up to the longer of: 17 years from the date of issuance or 20
    years from the date of filing.  So in this case, the patent would expire in 2015.

    I still stand by my argument that the specific implementation should be what the patent protects.  It's one thing if a company copies the implementation, it should be a different story if a company creates the feature oblivious to the existence of someone else's product.



  • @smbarbour said:

    I still stand by my argument that the specific implementation should be what the patent protects. It's one thing if a company copies the implementation, it should be a different story if a company creates the feature oblivious to the existence of someone else's product.
    That's just not a workable scheme. If access to patents is restricted, how could you avoid inadvertently copying someone's specific implementation? And if access to patents were kept open, how would you know if the creation of an identical feature was due to the existence of someone else's implementation? The most rational balance is to require all patents to be public, but eliminate independent creation as a defense.

    You can certainly argue that some other balance would be better. But to do it without mentioning how you would strike that balance or what other costs it would have is unreasonable.



  • @Huf Lungdung said:

    @joelkatz said:

    Ideas are not patentable. Only implementations of ideas (such as devices and methods) are.

    What about algorithms? Algorithms are ideas for how to solve a problem, whereas an imlpementation is a piece of code. What i4i recieved a pantent for was an idea.

     Algorythms are not patentable. The courts and/or congress have decided that in as many words. That is why you will never see the word 'algorythm' in a patent document. Pre Bilksi, the synonym used was 'Method'. Post Bilksi, some ricuculous circumlocutions like "Device consisting of a general purpose computer which, when executing certain code, becomes a special purpose computing device that ......"

     They are all algorythms, however, and the Office should toss them all out on sight, according to the law. But, as the Office relies on the corporations filing this junk for money, that's not going to happen.



  • @joelkatz said:

    @smbarbour said:

    I still stand by my argument that the specific implementation should be what the patent protects. It's one thing if a company copies the implementation, it should be a different story if a company creates the feature oblivious to the existence of someone else's product.
    That's just not a workable scheme. If access to patents is restricted, how could you avoid inadvertently copying someone's specific implementation? And if access to patents were kept open, how would you know if the creation of an identical feature was due to the existence of someone else's implementation? The most rational balance is to require all patents to be public, but eliminate independent creation as a defense.

    You can certainly argue that some other balance would be better. But to do it without mentioning how you would strike that balance or what other costs it would have is unreasonable.

    So the answer is that the first person to think up something has the exclusive rights to it and anything remotely similar for 20 years?  That doesn't sound like a reasonable balance at all.  If software patents are allowed (which they should in order to promote innovation), they should have to meet criteria to ensure that the patent is not overly broad or vague (such as the idea of putting metadata in another location for documents).  There should be a criteria to prove that this "innovation" is significantly different from other technologies and that it meets a need for a specific purpose.



  • @smbarbour said:

    So the answer is that the first person to think up something has the exclusive rights to it and anything remotely similar for 20 years?

    No. That's not how the patent system works. It's easy to find a small number of outliers to make it look that way, but they are the exception rather than the rule.

    That doesn't sound like a reasonable balance at all.  If software patents are allowed (which they should in order to promote innovation), they should have to meet criteria to ensure that the patent is not overly broad or vague (such as the idea of putting metadata in another location for documents).  There should be a criteria to prove that this "innovation" is significantly different from other technologies and that it meets a need for a specific purpose.
    There are rules exactly like that. Again, this patent likely would have been found invalid had Microsoft challenged it based on the right prior art early enough in the proceeding.

     In any event, your criteria are not needed. If it's not significantly different and doesn't meet needs for a specific purpose, then the patent is harmless. The patent won't cover anything sufficiently similar to anything prior to the patent, so there's no harm in patenting things that aren't significantly different. And if the patent doesn't meet any needs, then again the patent is harmless. There is no harm in someone preventing you from doing something useless.



  • @joelkatz said:

    My generic response to this kind of argument is this simple:

     

    OMFG, I'm actually not the only person on the planet who understands patents. We should, like, go out for a beer or something.



  • That's one way to fund your startup.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.