It’s no secret that I mirror my contributions to another site and plant them in this blog. If this bothers long-time readers, please let me know.

Mark Antony Kent submits the following take on Novell’s deal:

“Presumably this deal really goes back to Novell’s roots. Their heritage is entirely that of a proprietary software vendor, who’ve been competing with Microsoft for at least a couple of decades, and mostly losing.

As things got really bad for Novell, the board was very aware that the only growth area which was not dominated by Microsoft was in free software, so they looked around to see what they could do. The most promising of all the distributions at that time was Suse, based in Germany (important – not in the US), had good traction in Europe and was well respected globally.

Like most of the early distributors, Suse was doing financially okay, but no more than that, and was desperate for cash to fund their continued expansion; working in a commodity environment means working with very very narrow margins, having significant investment can make that much less painful.

So, it looks like a match made in heaven. Novell, the well respected US-based networking company, investing in, indeed buying, one of the big four Linux distributors. Everyone in both the stock-markets and the open-source world applauded the move; Novell had a new lease of life in the server arena, their traditional market place, /and/ got to push Suse for the desktop too. Until Ubuntu came along, Suse was considered by many to be the most mature business desktop offering.

Unfortunately for all concerned, although the match was reasonably successful, and Novell have enjoyed a steadily increasing penetration into the desktop and server markets previously dominated by Microsoft and Unix vendors, the rate wasn’t all that high, and was damaged by the SCO activity, something Novell found itself on the wrong side of, but was acutely aware it was in the spotlight.

At about the same time, the FSF and Prof Moglen began work on GPL3. GPL3 is intended to prevent the last great legal threat possibility against linux and free software, that of patent fud. When the GPL was first created, this wasn’t much of an issue, because nobody had been granting patents for trivial software, but, the USPTO was made a profit centre, and its managers realised very quickly that in order to make their profit, so that they could get their bonusses, they must grant as many patents as they possibly can; which is what they did. Patents have been granted for such triviata as an “icon on a screen” to such obvious developments with prior art, such as moving email on a packet-radio link (RIM suffered from that one, even though it’s been going on for decades).

So, as the SCO debacle reached its peak, the FSF people were considering how to counter this new threat against freedom, albeit a threat resulting from a series of blunders by the US Government, including starting to offer software patents, and then making money from granting them. Thus, GPL3 is born.

The evironment in which GPL3 came to pass should be considered carefully, and compared with GPL2. For GPL2, most developers were freelancing in some way or other, perhaps paid by some employer, but their work wasn’t germane to the company, perhaps students or academics, or enthusiasts working at night in their cellars. Thus, GPL2 addressed their fears very well – their work could not be taken by anyone else, improved and sold-on, without the improvements being returned to the code base. A reasonable proposition, although even then, major companies such as Microsoft were already calling it a Cancer, precisely because they were not able to take the work and pass it off as their own.

By the time GPL3 is being written, though, free software has become something of a victim of its own success. Many, if not most, developers are now working for commercial organisations, or, if not, at least they’re working for “not for profits”. Thus the whole attitude to patents is affected, if not dominated, by the legal departments of their employers. The concepts of “RAND” – reasonable and non-discriminatory, are preferred licensing methods by many companies, based on the sage advice of their senior legal people. Why? Because they still do not understand how software is developed, or why free software has been so successful. There is still the view that free software can be “de-commoditised” in some way, although any economist worth his salt will tell you that this is exceptionally difficult to achieve (but not impossible).

So, whilst the previous generation of academics, cellar-dwellers, students and part-time coders were happy to have their own work protected by GPL2, the current generation are having to look over their shoulder at their legal team for GPL3, but their legal teams do not necessarily understand why GPL2 exists; they might understand what it means, but they don’t understand why its there. Thus, GPL3 is getting a rough ride indeed from many people. An additional argument has been that GPL3 is not the right route to resolving the software patent problem, indeed, it’s just possible that this is true, however, pragmatically, it’s likely to be the only one which will work, unless someone seriously believes that the free software community can take on and beat the US government on patents granting – something I doubt very much.

Microsoft have learnt a lot from their funding of SCO to claim copyright on Linux. These lessons include that the approach was, in fact, surprisingly successful. There’s always a pundit or analyst who can make money from selling the FUD on in companies, and as most companies, and lawyers, are naturally very conservative and cautious, merely making a public accusation can be enough to stop customers from moving to linux. However, the second lesson has been that whilst the wrath of the free software community doesn’t scare them one bit, the steady loss of customers to the free software world scares them greatly. They have no fear of the FSF, of Linus T, of ESR or of RMS, but they are terrified of linux distributions which prove that free software is good enough for most uses, and they’re terrified of the GPL, which protects them.

So, they’ve gone for a second round of legal attack. This is nothing new for Microsoft, indeed, the SCO case was nothing new. From the beginning, Microsoft have been about manipulating the legal system to their own ends, and marketing heavily to persuade an unknowledgeable public of their wholesome aims. The agreement with Novell is just one more step from Bill Gate’s original letter regarding the copying of basic, sent just scant months after he’d stolen printouts himself in order to write the basic, Microsoft have never played this game with anything less than four Aces up each sleeve.

Microsoft saw GPL3 coming along like a high-speed train, and they were struggling to get over the level-crossing. They knew that they had just a few weeks, months at most, to get a patent attack in, before the free software world began to recognise the huge danger this poses, so, they did one of the things they’re best at. They charmed the board of a company – in this case, a competing company, but one which they knew was still cash-strapped, was not growing at the rate it wanted to, and was still painfully aware of how close it had come to litigation in the SCO case. Whilst IBM might have protected Novell, it might not have done, and where would that leave the boys from Utah? So Microsoft found very fertile territory in the Novell board. People whose real background was proprietary code, networking and unix; people who understood lock-in and licences, and people who were desperate for both more cash and for peace of mind that they would not be under attack from Microsoft. But, most of all, they were a company board who were desperate to do better than the market leader in their space, Red Hat, and also do better than the very well funded upstart, Ubuntu.

Microsoft played, as always, several Aces from each sleeve. They agreed to provide a seal of approval for Novell Linux; they agreed to stop fuding it in public. They wouldn’t actually stop, of course – they’d keep those Aces for later, and let Novell’s lawyers argue it out for as long as they could afford to, Microsoft never needs to keep any promises it makes, it’s rich enough to avoid that, another Ace up its sleeve. But more importantly, they agreed to not attack Novell’s customers legally, whilst agreeing not to reach a similar agreement with any other linux distributor.

Well, after a few hearty meals, glasses of beer, wine and spirits, how could an offer like that look anything other than gold-plated? What more could a CEO want than to protect both his company, and his customers, from Microsoft’s attacks? It would be bound to improve his company’s standing in the Linux world, and on the stock-markets. At least, that’s what a CEO, fresh out of the proprietary world would think. A CEO with experience of the free software world would’ve seen all the landmines, pitfalls and caltrops scattered along this road, and would also have been far more aware of GPL3 and why it was being written – to protect his company and his customers from precisely this problem.

I suspect that the real winner from this debacle with be GPL3, as it will highlight the persistent danger which Microsoft poses to the free software world. Microsoft have certainly lost the server battle and the mobility battle and are in the process of losing the embedded battle and the console battles; they look like they will also lose the desktop battle, although it might take longer, but they are very very cash-rich, very powerful, and have far from abandoned the war.

I strongly recommend that anyone who’s written off GPL3 as unecessary reconsiders their position in the light of the Novell/Microsoft deal.”

This entry was posted on Thursday, November 23rd, 2006 at 3:19 pm and is filed under Novell. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.