A few years ago I bought my first house. The process was overwhelming at first, but due to the assistance of an attorney, a mortgage lender, an appraiser, an assessor, a surveyor, two insurance agents (one for the house and one for the title), a house inspector, a real estate agent, etc., the purchase occurred without problem.
But why were so many people involved? Why so much complexity?
Each, aside from having a professional specialty, looks out for a specific interest and has a duty to a particular participant in the transaction: to the buyer, to the seller, to the lender or to the tax collector.
In the end, I have my house, my land, and a little piece of paper called the Quitclaim Deed, which conveyed the seller’s interest in the property to me. The deed specifies the parties to the transaction, the amount paid, and references a legal description of the property, which reads in part like:
N 20-07-00W 94.41 feet to a lead pipe, N 21-06-30W 372.04 feet to a stone bound, N 63-17-40E 291.05 feet to a stone bound, 52-29-20E 360.60 feet to a stone bound, etc.
This style of land description, called Metes and Bounds, is particular to New England, and has been in use here since colonial times.
What is interesting, in terms of the real estate transaction, is the division of responsibilities. The attorney looks at the paperwork, but the surveyor verifies the property description. The attorney ensures that the form of the deed is in accordance with local law and customs, but he is not going to be able to tell you that your garage has been built half on your neighbor’ plot. That is the job of the surveyor.
In the end, I am happy only if I have successfully purchased the property I intended to purchase. If the formal paperwork is executed properly, and if the survey matches the bounds of the property I believe I am purchasing, then I am happy. If either of these fail, I will not be pleased with the transaction, even if the other criterion is met.
I am reminded of the above mechanisms when thinking about IP issues in Microsoft’s OOXML. The analogy works like this:
You only have full rights to implement OOXML only if you are satisfied with:
- The Conveyance = The formal language of Microsoft’s Open Specification Promise is bullet-proof
- The Survey = The technical detail of the OOXML text is complete and accurate and matches what you need to implement in your software
- The Title = Microsoft owns (and continues to own) all of the patents required to implement the portion of OOXML you wish to implement in your software
Much has been said about the first item. Microsoft has given numerous assurances, commissioned reports from very expensive law firms that all affirm that the formal language of Microsoft’s Open Specification Promise is air tight. Microsoft’s Jason Matusow boldly proclaims: “There are NO intellectual property rights issues with Open XML“. Jason claims that Ecma (who failed to notice thousands of technical flaws in OOXML before approving it) and ISO (which has been in a “see no evil, hear no evil” state of denial for the last year) both “Have Publicly Declared that No IPR Issues Exist”.
I guess I’m just not as easily impressed. Who there is looking out for your interests? Who has a fiduciary duty to you? Not Microsoft. Not Ecma. Not ISO. So you better watch out for your own interests.
So let’s ask, what is Microsoft actually promising? The Open Specification Promise says in part:
Microsoft irrevocably promises not to assert any Microsoft Necessary Claims against you for making, using, selling, offering for sale, importing or distributing any implementation to the extent it conforms to a Covered Specification…
…“Microsoft Necessary Claims” are those claims of Microsoft-owned or Microsoft-controlled patents that are necessary to implement only the required portions of the Covered Specification that are described in detail and not merely referenced in such Specification.
Certain rights are given to you, and the bounds of these rights are circumscribed by the stated restrictions. In this case, instead of enumerating exactly what patents are made available, they are described implicitly by these criteria:
- Microsoft owns or controls the patents
- They are necessary to implement “only” the required portions of the OOXML specification
- That portion of the specification must be described in detail and not merely referenced.
I think of the above as the “Metes and Bounds” of Microsoft’s OSP. Finer minds than mine have tried to make sense of this, parsing the language, thinking about how every word could be interpreted, etc. But you can pour over this all you want, and not determine what your rights are. Edward Coke, Lycurgus of Sparta and Moses could all declare this to be the finest legal document since the Code of Hammurabi, but that would mean little.
The key point is that because the way the OSP is crafted, your rights are intrinsically tied to the quality of the underlying OOXML specification. If that text is vague, and lacking in precision, then your rights are reduced. Your rights are secured only to the extent the text is detailed. Low quality, low level of detail or missing detail equates to lack of protection for the implementor.
Since no one who is pontificating on how there are “no IP issues” has actually read the OOXML standard, I think they should be a little less effusive in their praise. But then again, they have no duty to watch out for your best interests, have they?
Let’s look at an example, to make this concrete. DIS 29500 contains the following passage which we’ve discussed before:
126.96.36.199 footnoteLayoutLikeWW8 (Emulate Word 6.x/95/97 Footnote Placement)
This element specifies that applications shall emulate the behavior of a previously existing word processing application (Microsoft Word 6.x/95/97) when determining the placement of the contents of footnotes relative to the page on which the footnote reference occurs. This emulation typically involves some and/or all of the footnote being inappropriately placed on the page following the footnote reference.
[Guidance: To faithfully replicate this behavior, applications must imitate the behavior of that application, which involves many possible behaviors and cannot be faithfully placed into narrative for this Office Open XML Standard. If applications wish to match this behavior, they must utilize and duplicate the output of those applications. It is recommended that applications not intentionally replicate this behavior as it was deprecated due to issues with its output, and is maintained only for compatibility with existing documents from that application. end guidance]
I don’t think there is anyone out there who would argue that this part of the specification is “described in detail”. In fact it disclaims all detail. Further, it would be hard to argue that this portion is required, when the text itself so clearly recommends that applications do not try to implement it. So, I wonder whether Microsoft’s Open Specification Promise would apply to any patents necessary to implement this feature.
Would anyone assert that a plain reading of the OSP suggests otherwise?
But you might say, “Please Rob, you can’t be serious. Who would try to get a patent on laying out a footnote? That just doesn’t happen in the real world.”
But consider for Microsoft’s patent application, “Method and computer readable medium for laying out footnotes” (US20060156225A1). I’m not saying that application matches the above feature in the standard, but if it did, if it did, I’m not so sure that the Open Specification Promise would cover.
This is the essential risk with OOXML, that its rushed preparation has lead to a 6,000 page standard full of errors, omissions and ambiguities, and your rights as a implementor are reduced by this low level of quality. This is not necessarily from malice, but from neglect.
Now what if I told you, that Ecma, as part of its proposed Disposition of Comments report has added a detailed functional description to the footnoteLayoutLikeWW8 element. This is a good thing, right? Not only do you have more technical detail, but you also have more IP rights because of that detail.
So this should give us pause. An aggressive review of OOXML (thank you) has resulted in implementors having more rights under the Open Specification Promise. But along with that joy should come concerns that the OOXML standard has been very sparsely reviewed. Many NB’s complained about not being able to review it thoroughly. I know I have not read more than 20% of it. So the full set of features which are unimplementable due to vague or missing detail is unknown. This same set of features may require Microsoft patents which are not eligible under the Open Specification Promise for lack of such detail.
So this is yet another reason why a rushed review is so harmful. Not only does it leave OOXML full of technical errors, including portability, accessibility, and interoperability flaws, but the resulting pervasive lack of detail means that implementors have far less IP protection than they may think they have.
You are buying a house. The attorney has blessed the paper work, the title insurance is in hand. The surveyor shows up at the closing and says, “I’m sorry. Your property was so big and complicated, that we were able to only survey 20% of your property line in time for the closing.”
The lawyer grins and says, “All the paper work is in order. All you need to do is sign.”
What do you do?
Chris Ward says
It’s different in the UK.
Back in 1066, the place got invaded by William the Conqueror. He declared himself king, and thereby granted himself a monopoly on the land.
It’s never been conquered since. Things have got close a few times, but the Kingdom still stands. And now all the land (from high tide mark upwards) belongs to Queen Elizabeth. And we rather assume that the next monarch will be King Charles.
As loyal subjects, we hold the land on which our houses stand from her, and she guarantees it. She’s got a list, in places like http://www.landreg.gov.uk/ .
So, generally no ‘quitclaim deeds’. No title insurance. When you buy and sell land, you pay your stamp duty (taxation without representation) and the monarch of the day registers the new title.
Apart from that, the process is very similar to that in the US. Professionals who need paying for professional services.
Chris Ward says
Oh, to further the analogy.
From the high tide mark down, no British monarch has ever given anyone the exclusive right to occupy. No-one can hold UK land below high tide. It still belongs to the Crown, everywhere around the coastline.
Except for the odd Royal castle, or the occasional Ministry Of Defence firing range, we treat it as having a General Public Licence. Come and go as you please. Anywhere on it, every beach in the Kingdom, no charge. We hold it in common, we share.
Luc Bollen says
Here is a quite interesting view and comment about the OOXML Open Specification Promise :
Obviously, according to Microsoft itself, the possibility to use the GPL licence with OOXML is… not obvious.
Chris Ward says
About the OSP and Brendan Scott’s comment — well, what did you expect them to say ?
Something similar happens with IBM Lotus Symphony . Obviously IBM would like people to take it and use it; that’s why it’s there, and why its price is zero. And IBM doesn’t mind if you give — or sell — copies to your mates.
But will anyone else mind ? There are (for example) laws in various countries of the world governing possession, use, and export of technology which can be used for encryption of data. In most of the Western world, we tend to think it’s necessary for ordinary commercial operations and for exercise of personal freedoms; but some countries take different views. And Symphony might in some circumstances encrypt something.
So how should IBM know (or why should IBM care) what your local laws are, or whether you have an exemption ? You have to seek your own advice on that one, if it matters.
6 billion copies of IBM Lotus Symphony, one in the hand of every person on the planet, here we come.
OOXML ? Like typewriters. We used to need them.
Microsoft might have said something like:
“The Open Specification Promise is consistent with most FOSS licenses including the GPL. Please understand this is not a legal advice. For anything that matters, please consult with your lawyer.”
Of course since presumably the OSP is not consistent with the GPL, they cannot say that.
Doug Webb says
Many of the problems and complications with OOXML arise around fidelity of writing deprecated formats. I don’t understand why one would need to write documents in deprecated formats. Wouldn’t it be far easier to convert old documents to XML going forward? The only conceivable need is when one wants to work with someone who has only an old version of Windows and an old version of Word. In this case wouldn’t it make far more sense to provide a tool to back-convert an XML document to these old formats instead of burdening the XML format with all the old mistakes? If you wanted to build a car using the OOXML design principles it would have a hydrogen fuel cell, a 2 stroke gasoline engine, a 4 stroke gasoline engine and a deisel engine as well. Sure, you’d be able to fill it up anywhere, but would it really be worth dragging all the extra weight around?
@Doug, that’s a good analogy. Consider that Word 2007 contains with in its code, modules that can deal with these old legacy formatting options. So when creating an XML format they had two choices:
1) The lazy choice, where they just indicate the legacy choices as flags in the OOXML format. This requires zero effort for Microsoft and makes the documents harder to consumer for competitors.
2) Convert the legacy settings into the preferred markup now. This would require that Microsoft make a greater investment in their conversion code and would make OOXML documents more accessible to Microsoft competitors.
So what does Microsoft do? Yup, the anti-competitive choice. Anyone surprised?
Chris Ward says
So how does the competitive landscape stand ?
Presumably IBM thought that OS/2 Warp 4 was the end of the line for ‘commercial operating systems for personal computers’. Competition hotted up, Amazon do a lovely line in Linuxes, and it doesn’t really take IBM skills to place an order there.
Now, it’s Lotus SmartSuite’s turn. ‘Commercial standalone office productivity software’ can be bought here, but not many people do. And the chance that it will be developed to support either ISO26300 or DIS29500 looks rather remote. In fact if you search for SmartSuite on the IBM direct web site, you get a banner ad for IBM Lotus Symphony pointing out that it’s available by direct download from IBM, at no charge.
Moving up the value chain, IBM Lotus Notes is a pretty good bet if you’re a large business or government wanting a well-controlled collaboration environment for your employees, and it’s selling like hot cakes. Bringing in giga-dollars.
So, the theoretically-anticompetitive behaviour has knocked out the market for OS/2 and SmartSuite (the hypothetical competitors to Windows and Office); but they have been replaced by items which are standards-compliant and available to all at no charge.
Such is the wonder of the participatory Web 2.0 . We share our problems and our solutions, and eventually we get there. For some businesses, it’s good.
Where next, though ?
Mr. Ward, I consistently read your comments at OOXML discussion sites. And I have to say, after several months of this…
“Truly, you have a dizzying intellect.”
I don’t believe I’ve ever understood a single comment you’ve made.
It *is* my supposition that you work for IBM, probably in the Lotus division or with Lotus technologies in some way. And that is a collection of technologies about which I have a substantial knowledge, but I’m afraid that, to me, you are Jeremiah the Bullfrog.
Chris, I think you are oversimplifying a little. Even in England, land may be owned by The Crown or one of two Dukes (Cornwall and Lancaster). Then there’s Scotland, which is different, and within Scotland there are Shetland and Orkney, which are different still.
The UK is certainly different, but your implication that it’s simpler may be misplaced…
Are you suggesting that allthough you are apperantly allowed to recreate normal Office document behaviour defined in the OOXML because of the OSP that you would be unable to use the compatibility items because Micrsoft could have tyhe technology for those deprecated compatibility items (essentially bugs in older Office versions) patented and would be able to use those patents when you use the deprecated compatibility items.
Are you this desperate that you are now stating Microsoft has patented their old bugs and are oging to use those against a party after they have submitted them to ISO?
Is that how IBM becomes the yearly leader in patentclaims, by claiming their bugs as patented technology ?
Chris Ward says
Fair enough. I work for IBM, but I don’t represent IBM. At the moment, I’m trying to figure out how to exploit the tens-of-thousands-of-processors in a BlueGene; if you like, spending IBM’s research dollars, and trying to move things on to where it’s attractive for clients to invest their research dollars.
On that scale, ODF and OOXML is a bit of a sideshow. It’s going to come to a vote; like your Presidential election, one side is going to win and the other is going to lose; and we’re all going to have to live with the outcome.
I think what we’re trying to do here — and Rob, it’s your ‘blog’, correct me if I’m wrong — is shed enough light so that the various participants can see what they are letting themselves in for, and choose wisely.
When I think about IBM’s position, it is most easily understood by looking at who it sells to, and what it sells. The ‘who’ tends to be ‘governments and large businesses’, and the ‘what’ is ‘hardware, software, and services’. IBM will sell to other organisations and individuals, of course; the ‘direct channel’ does a lot of business, and a fair bit is sold through partners. But the sales force tends to find it most worthwhile concentrating on governments and large businesses.
And of course IBM tries to sell them what IBM thinks they want to buy; as a government or large business, depending on an open standard puts you in charge of your own destiny.
Some lines of business eventually commoditise. IBM is a high-innovation relatively-high-cost full-service supplier; and if IBM can’t improve a product by innovation, IBM has to either sell the product at a loss or exit the business. Selling the product at a loss won’t bankrupt IBM, but is commercially pointless and tends to cause competitors to complain, so IBM tries not to do it.
The large case recently was with Personal Computers; if you were an IBM Personal Computer salesman, there was a fork in the road. You could either remain an IBM salesman and retrain to sell something else, or you could transfer to Lenovo and become a Lenovo Personal Computer salesman. There is now no such thing as an IBM Personal Computer, and therefore no such person as an IBM Personal Computer Salesman.
This re-slicing of the business is ongoing. OS/2, the IBM operating system for Personal Computers, was a sweet thing; worked very well; the only problem was that no-one would buy it. So IBM sold that one off, too. If you really want, you can get one from Serenity Systems; but not from IBM.
Lotus SmartSuite, broadly speaking the IBM competitor to Microsoft Office, looks to be going the same way; particularly if IBM is going to tell people that they can go get IBM Lotus Symphony at no charge. We view it (I think) as a form of advertising for Lotus Notes. It’s rather like what happened when ATT stopped making money in the handset rental business, and moved into the ‘infrastructure’ business; a revolutionary disruptive change at the time, but looking back it was an obvious progression. Why fight the OpenOffice tide, when you can accept it and build on it ? Not all software developers hang out in corporations; universities develop software (for rather different reasons), they eventually close down whatever lead the corporations have, and if the universities are to train the future technologists for the corporate workforce, their work needs to be encouraged rather than suppressed.
Anyway, to OOXML and ODF. To me, OOXML looks to be a component of a solution which only is deliverable in context of Microsoft Windows and Microsoft Office. Can you imagine anyone going to the trouble of implementing OOXML for a Sony Playstation ? Whereas ODF already has been. And expecting IBM to invest in developing SmartSuite to the point where it can digest OOXML documents is rather like expecting ATT to invest in developing telephone handsets; both IBM and ATT have more productive things to invest in nowadays.
The 2 forms of XML seem to be for fundamentally different purposes. OOXML is a documentation of the behaviour of one vendor’s product; ODF is the result of the deliberation of many stakeholders … vendors, purchasers, academics … of how they would like things to be.
There are huge commercial forces at stake, of course. Microsoft could spend a billion dollars to attempt to swing the vote their way; IBM likewise. You might attempt to justify it as spending a billion dollars to buy good players in an attempt to win the Superbowl. But spending a billion dollars like that wouldn’t be good for ‘the world’, and I’m sure it takes great forbearance on all sides to only do what is within the bounds of commercially acceptable behaviour.
On towards the BRM, then.
@anonymous, the legacy compatibility flag was merely an illustration. The risk is much broader and reaches to every feature of OOXML that is poorly defined, lacks detail or description.
In other words, this is a pervasive problem.
Chris, thank you for the dramatically more understandable post. :-)
Mike Rhodin has been open about the fact that Symphony has been built with the intent of commoditizing productivity software, by the way. I don’t think anyone could reasonably think that’s a hidden intent. And, as you observe, commoditizing it threatens a massive revenue stream for Microsoft, while helping software users enormously.
[quote]Of course since presumably the OSP is not consistent with the GPL[/quote]
The OSP is not related to the GPL.
GPL deals with IP rights of contributors to the OSS software and does not grant 3rd party rights.
The OSP has to do with Microsoft patent rights which are in effect 3rd party rights.
So the rights granted trough the OSP therefore are additional rights to the rights granted by GPL licensing.
This is simular to the covenant not to sue by Sun (and likely a similar promise by IBM/Lotus)
Thus the OSP or CNS promises are compatible with all OSS licenses because it grants extra 3rd party patent rights.
Oh…Hal…would you mean that you have a fiduciary duty to the rest of us?
Why should we trust that you, as a microsoft partner, have our rights as the first concern when you answer?
You claim that the OSP is GPL compatible…sorry but that counts for nothing unless it is Microsoft that make the promise.
Microsoft that by all means know what they plan to do with the OOXML patents refuse to tell us if they will use these against GPL software or not. It is not a matter that the rest of us can solve by consulting lawyers, Microsoft need to tell us what their true intentions are because the hold the ball.
A personal speculation of mine is that what Microsoft intend to do is to sell the patents to patent trolls later and make these try to stop the GPL momentum.
Fortunately I don’t think it will be enough to stop the GPL, but meanwhile you could at least spare us from hearing empty speculations about that Microsoft perhaps having good motives for not explictly allowing GPL code to implement OOXML.