Hmmm, if you think so, perhaps you'd better read up on this, then

A call to tailor the patent system to special needs of the "IT industry" would be a dead end, a nice way to ensure that the opposition against software patents stays irrelevant. A major achievement of FFII has been to get people away from such dead ends.

Thus your predictions and those of Mark don't necessarily contradict each other.

Some sidenotes: you write that the CAFC has been "systematically dismantling all barriers to patent quality". I wish they had actually been doing that!

Also, I don't understand what you mean by a "joining the rest of the IT industry to lobby for … a special patent regime for software". Certainly FFII doesn't want any "patent regime for software", neither a normal nor a special one, and I don't see the "IT industry" in need of that special regime either. The meaning of the word "patent" is quite narrowly determined by convention and by international treaties, there is practically no room for maneuver. A "patent" will, for the time being, always be a broad monopoly right, defined by written claims, determined and validated through expensive judicial procedures. There is no way that this concept can be made to work for the "IT industry". If anybody is lobbying for a "special patent regime" at the moment, then it is Microsoft, through their reform proposals to the US Congress, which won't change much, if they are feasible at all.

Today, right now, the software industry is divided and it is at war. The battle lines are drawn, almost every firm in the industry is clearly on one side or the other. The fight is over a simple thing - how to survive in a changing world.

I wrote in 1999 about “The Ice Wars”, in which the Internet revolution turned software from a precious commodity into a raw material for vast new industries. I said then that there were those firms who looked forwards, and those who looked back. Those who wanted to build new empires on cheap ice, and those who wanted to sell ice by the pound. Those who understood the future, and those who denied it.

In 2007, the picture is even clearer, and many firms that sat in the middle of the field in 1999 - like Sun, Oracle, and Nokia - have placed, or are placing their standard alongside IBM, Cisco, and the whole free/open source software community focused around GNU/Linux.

The only major software firm still fighting the inevitable future is Microsoft, and it is fighting so viciously, so aggressively, and doing so much damage in the process, that “threat” is an apt term. The fact that Microsoft is doing harm to itself does not lessen the harm it is doing to others.

Let me take two examples. First, software patents, and second, word processing.

When we (the FFII, of which I'm President) fight against software patents in Europe, we see who is on the other side. Lobby firms like the BSA, EICTA, ESA, and Comptia can only hide so much. In 2005, large EU telecoms and software firms, and Microsoft, were trying to buy a software patents law. They were helped by the patent industry (patent lawyers and the European Patent Office), who profit from any expansion in the patent system. Today, most of the European firms have dropped out, except for SAP and Siemens, and Microsoft still feeds the patent industry and still lobbies for software patents in Europe, from its Irish tax haven.

Siemens and SAP want software patents because they fear competition. Microsoft wants them for the same reason. It has watched free software grow from a curiosity into an inevitability, from a geek hobby into the mainstream. It knows that Linux is the future, the TCP/IP of operating systems. It sees every one of its market control tools being circumvented. It sees its own death, for Microsoft is like the Roman Empire - no compromise, no half-share. It's everything, or it's nothing, and Microsoft is counting on patents to save it.

Nokia, who wrote in 2004 that “all of Europe's innovators, including individual inventors, small and medium size enterprises (SMEs), as well as large multinational companies, require patents to protect their inventions, provide incentives to undertake research and development in Europe, and to promote licensing and technology transfer” today says that “this whole discussion might have an impact on 3G technology”. It's paying half a billion dollars a year to Qualcomm. Half a billion! Every Nokia handset carries a Qualcomm tax. And that's just one case. Nokia has patent license deals with dozens of firms.

Software patents, ultimately, are a tax on all firms that make products, and a tax on all consumers. The economics of patent risk (sue but avoid being sued) drive firms like Qualcomm, and Thomson, to become pure “IP” firms, which means abandoning the market, and cashing in on patent portfolios. It's short-term logic and long-term suicide: a firm that makes nothing cannot innovate. These firms are basically retired, and when their patents run out, will be worthless.

Microsoft, a market-oriented firm if ever there was one, should be as hostile to software patents as any free software developer. As Mark Shuttleworth said, “within a few years, Microsoft themselves will be strong advocates against software patents”. I've said this to my FFII colleagues often. Logically, Microsoft should become a member of the FFII and make huge donations to help us fight the patent industry's grip on technology! But large firms do not always act logically.

Instead, Microsoft thinks it can control the patent industry and exploit it. Here is what I think Microsoft is doing. First, it's understood where free software is going. It took them a while but with every Tom, Dick and Harry moving as fast as is decent to leverage the community, even the most conservative bean counter is going to ask, “if everyone else is doing it, why aren't we?”

But second, unlike the underdogs who are relishing the prospect of competing against Microsoft on what becomes a much more level playing field, Microsoft does not want openness, community, or freedom, unless it's strongly branded, properly licensed, and seriously profitable. Microsoft sees software patents as the key to this, and it's willing to take a great deal of pain to play the game, because it sees no alternative. It knows that it's a fat sitting target, but it's gambling that the market that Linux & Co. underpins is even fatter.

So how to 0wn Linux? Simple: divide and conquer. Start by taking over the commercial Linux vendors. Build an “IP bridge” that “indemnifies” clients of one vendor for “infringements” of “patents”. Spend a lot of money to promote this vendor's Linux. Watch the other vendors lose market share. Watch them resist. Keep pumping money into the “approved” Linux. Avoid litigation that could dispel the mists of FUD. “We'll never sue developers [but watch out, users!]”. Watch the other vendors come back to the table. Now turn the screw.

Microsoft may have its eyes on Novell's Suse as their “MS-Linux” but I think they see all commercial Linux distributions as fair game. Microsoft never settles for part of a market: they always play for first place. Once the top three or four commercial vendors have bowed to the inevitable (so goes the logic in Redmond), the Linux community will fragment and all the money will start to move towards the safe, integrated, supported products.

And thus Microsoft will build its MS-Linux, not by repacking Debian, but by carving out a licensing empire with Linux vendors. Microsoft has been at the receiving end of such license deals often enough - after years of paying for MP3 licenses from one consortium, it found itself owing $1.5bn more to a second group. Of course Microsoft thinks of this as a nice business, and wants to be at the collecting end, for once.

Redmond's strategy - if my idle speculation is accurate - depends on some unproven assumptions: that no-one will fight back in the courts, for fear of starting a war of mutually-assured destruction; that Microsoft will eventually earn more than they pay out in patent claims; that Microsoft can assert its patents in all three major economies (US, EU, and Japan); and that the patent industry can be kept under control. History will tell whether those assumptions are accurate.

Europe is an important battle-ground because it does not officially grant software patents at all. Europe's exports account for 52% of the global trade in services, and 44% in products, compared to 17% and 15% for North America. If Microsoft's ambush on Linux fails in Europe, it will collapse globally.

But in its single-minded campaign to get the software patents it needs to own Linux, Microsoft is enabling a much more dangerous enemy: those specialised firms that do nothing except buy the rights to what they think are good ideas, so they can hijack future markets. Patent trolls or “non-product entities” may consist of one lawyer, or hundreds, and may have only one patent, or thousands. What they all share is a focus on earning as much as possible from product-making firms. The price of a patent has nothing to do with the “technology” it's providing, and everything to do with the size of the customers' wallet.

Patent trolls are not “evil”, they are simply working a system. But so is the malaria parasite. That does not make it any less harmful. Intent is irrelevant, it's outcomes that count. And parties who profit from patents are dangerous because they get rich, powerful and lobby hard for more patents and weaker patent law. It was patent trolls, patent attorneys, and patent bureaucrats that turned the US patent system into a circus by taking control of the Circuit of Appeals of the Federal Circuit (CAFC) and systematically dismantling every barrier to lower quality.

As it happened in the US, it's happening in Europe. The patent industry has achieved much the same at the EPO's Technical Boards of Appeal and - with Microsoft's firm backing - is lobbying for a single EU-wide patent court that it can take over and run, as it does CAFC. Can anyone trust or control the patent industry?

The IT industries of the US, EU, and Japan have much to fear from expansion of the patent system. Once created, these patent-only firms will fight tooth and nail against any reform of the patent system. Their 20-year grip will cause entire fields of innovation to wither and dry up. People stop investing in areas that are heavily patented, as Nokia have discovered. And this leaves huge holes for Chinese, Taiwanese and Korean firms - who have a huge and mainly patent-free Asian market - to innovate and take intellectual ownership of much of the IT sector.

Let me make a prediction: in ten years' time, most new standards in fields that are heavily patented - audio, video, telecoms, etc. - will be Chinese. US and EU firms will become clients of Asian “intellectual property”, just as Asia has to pay patent taxes to sell its goods in the US and EU today. Popular and political support for the patent system in the US and EU will collapse, but with economic power shifting towards Asia, it won't matter. It'll take yet another cycle, twenty more years, before the global economy finally shakes off the parasite that the US/EU/JP patent system has become.

So Microsoft, by feeding the trolls, is setting itself up for another thirty years of billions-per-year extortion. If it holds the course that long. What will probably happen is that as the advantage starts to shift to Asia, and Microsoft sees its Linux strategy failing, it will, as Mark Shuttleworth said, go through a 180-degree turn, switch off the evilness, and join the rest of the IT industry in lobbying for open standards and a special patent regime for software (which may mean anything from 'no patents on software' to new forms of 'patent' that promote disclosure without harming the common good). Will Asia accept? Not if it means losing a profitable licensing business.

Now, I mentioned standards. In the traditional IT universe, firms compete to get “their” technology into standards, so they can join the patent pool that collects, if the standard is a hit. But in the internet universe, firms compete to contribute to patent-free open standards, which become the basis for whole new economies. There is no question of what works better. Open standards like SMTP, TCP/IP, HTTP are orders of magnitude more successful - in terms of creating new markets - than even the most successful patented standards, like MP3, and GSM.

And one of the banners that divides the two sides in the IT industry war I spoke of is their attitude towards standards. The reactionaries see standards as tools to control markets, and the progressives see standards as tools to create markets. Nowhere is it more clear than in the standardisation of word processing documents.

ISO - the International Organization for Standardization - has a standard for word processing documents: ODF, or Open Document Format, ISO/IEC 26300:2006. Like any good standard, ODF is built by consensus, and widely implemented. Vendors are discarding their proprietary formats and switching over to ODF. IBM's Lotus Notes will switch to ODF in its next release. Almost all word processor products have already done so. Except one: Microsoft Office.

And Microsoft sees ODF as a direct challenge to its hegemony. And its response? Unbelievably, to push to get its own format in place as an alternative ISO standard. “Office Open XML” as it calls it, is also an XML format, but is a single-vendor standard protected by patents and secrets, like any old proprietary standard.

It's one thing to hear Microsoft argue that two standards create more choice. It's quite another to see the way they are getting OOXML accepted by ISO.

ISO, like most standards organisations, depends on good will and ethics. ISO processes have never had to deal with a hostile take-over attempt, up to now. Microsoft has launched a wide and aggressive campaign to stuff the relevant ISO committees with its friends, and to sabotage the delicate process of meetings and reviews so that OOXML gets through unchallenged. Its antics are going largely unreported.

When Microsoft has won its fight to have OOXML labeled “ISO standard” - which I think it may, even though the community is fighting back - what damage will it have done to ISO? I fear that ISO will take a long time to recover. The network of contributors and volunteers who work on so many vital standards will become disgusted and will give up. The sight of a monopolist buying and bullying a respected international standards organisation into signing on the dotted line will devalue all future ISO work. And without a working standards process, how does the IT industry continue to move forwards?

Even though it's not hitting the news right now, Microsoft's interference with the ISO process will not go undocumented, and ISO rubber-stamping of its proprietary standard will only delay the inevitable. There will eventually be a single document standard, it will be ODF, and Microsoft will eventually implement it, but in the meantime we're all going to pay the price once again.

I've discussed two cases where Microsoft is trying to change the rules to suit itself. In both cases, as in many others, it will fail but in the meantime waste a huge amount of other people's time and money. The many volunteers who make the FFII live would be able to go back to their jobs and families if Microsoft just faced reality instead of trying to bend it. Microsoft is big, powerful, and brutal. It has killer instincts. It has patience, it can detect weakness from a far distance, and it does not let go, once it has its jaws around someone's throat. It's really like a huge shark. And like a huge shark, it has a tiny, prehistoric, stupid brain.

So yes, I'd argue that Microsoft is most definitely the threat. They hold the sword at the throat of much of the IT industry. When Microsoft understands, finally, that the future belongs to the good guys, that “do no evil” is more powerful marketing than “do more evil”, that in order to have a future, a firm must not hold onto the past… then, finally, we can say that Microsoft is no longer a threat, and we can turn our full attention to the trolls and the lawyers and the bureaucrats.