NetApp tilts at windmills – and ZFS

by Robin Harris on Friday, 7 September, 2007

Or maybe shoots self in foot
NetApp announced on Wednesday that it was suing Sun for patent infringement for ZFS. Sun immediately announced it would indemnify customers against NetApp’s claims, so there is no reason for, let’s say, Apple to stop work on ZFS.

I’m a fan of ZFS, Sun’s open-sourcing of it and of Thumper, the ZFS-based storage system. Could a cold-hearted NetApp really strangle baby ZFS in its crib? Nah!

In fact, there is a better than even chance that this suit will turn out to be a Very Good Thing for the software business, customers and OSS. And a Bad Thing for patent lawyers other than the ones billing $600/hr to NetApp and Sun.

Can’t make an omelette with breaking some eggs
Or fix the US patent system with breaking some precedents. Which is what the US Supreme Court did in their unanimous decision in KSR v. Teleflex. KSR requires a deep rethinking of the nature of invention in a network world. The NetApp suit will be one of those spurring much-needed change.

He said, she said
NetApp says that Sun attacked it using STK patents. Sun says NetApp originally approached them asking to license the patents. NetApp has angered at least one customer enough to withdraw their reference – a request made on Dave Hitz’s blog.

This started when STK – before acquisition by Sun – asserted patent infringement against NetApp. STK was being “dressed” for sale by slashing R&D while scrambling for every bit of revenue and potential revenue that might be valuable to an acquirer.

Dave took some pains to differentiate their suit from the Microsoft-financed attack on Linux by SCO, as well he might. A warm and generous NetApp, secure in its leadership, might have donated the patents in question to the OSS community, as IBM has, but I don’t hold it against them that they didn’t. It would have been cheaper to donate them and better PR, but as long as we an overworked and underpaid USPTO issuing patents, we have to expect that people may sue over them.

Mindless optimism or too many 3-2-1 Margaritas?
In the Teleflex case the Supremes upended several decades worth of process for determining the non-obviousness of an invention without “hindsight bias.” Inventions have to be non-obvious because otherwise it isn’t much of an invention and doesn’t merit legal protection. But figuring out how “obvious” an invention was at the time of invention, especially years later when it may be in wide use, isn’t easy.

The courts have developed the TSM (teaching, suggestion or motivation) test, under which an invention “. . . is only proved obvious if the prior art, the problem’s nature, or the knowledge of a person having ordinary skill in the art reveals some motivation or suggestion to combine the prior art teachings.”

For example, a lead-acid battery uses a liquid electolyte and two metals to generate electricity. So does that make any battery design using a liquid and two metals obvious?

If it is a fruit-powered clock that uses established principles, no. But if the battery uses a new electrolyte and materials that those skilled in the art generally rejected, then yes.

The key point in the Supreme’s decision is

The diversity of inventive pursuits and of modern technology counsels against confining the obviousness analysis by a formalistic conception of the words teaching, suggestion, and motivation, or by overemphasizing the importance of published articles and the explicit content of issued patents. In many fields there may be little discussion of obvious techniques or combinations, and market demand, rather than scientific literature, may often drive design trends. Granting patent protection to advances that would occur in the ordinary course without real innovation retards progress and may, for patents combining previously known elements, deprive prior inventions of their value or utility.

[bolding mine]

Although the court left the details undefined, which will force smart lawyers and judges to think these issues through, this is a much tougher standard than the courts have been using. And when the NetApp patents are examined in that light much, if not all, of the patent claims will be voided. Not that NetApp will take it that far.

Is solving a problem an invention?
For example, NetApp makes some claims about Sun’s use of their snapshot copy patent. Lampson and Sproull’s file system for the Alto computer – forerunner to the Mac – did snapshot copies in 1979. So when an expert engineer is solving a customer problem and puts together known concepts – like snapshot copy – is he inventing?

Under the standard the Supreme Court put forth the answer is no. Combining existing technologies in a new solution is not an invention if the combination would be likely to occur to anyone skilled in the art.

The StorageMojo take
I think this case will turn out to be a watershed event in dismantling our broken software patent process. The Supreme Court’s unanimous decision is the equivalent of hitting reset on the patent system.

Software invention has been fast and furious for the last 60 years with many brilliant people contributing ideas freely. These smart people were able to imagine many possibilities that, due to technical limitations, could not be implemented at the time.

Moore’s Law and network economics keep changing the playing field so old ideas often becomes new again. With the Internet those ideas are more readily available than ever before, accelerating the rate at which ideas can be combined and developed.

It will take years for new case law based on the KSR case to develop, but I’m optimistic that when the dust settles we’ll have a system that rewards the hard work of unique invention while protecting an engineer’s right to use existing ideas to solve hard problems.

Update: Unbeknownst to me, the House passed a bill yesterday that reforms some of the draconian parts of the patent system such as assessing damages based on the total value of the product rather than the infringing component.

The problem of software patents isn’t the same as the hardware patents. It appears to me that there are very few new concepts in software and that most of those are developed by the academic community. When engineers combine existing concepts to solve a new problem – such as Amazon’s 1-click technique, the KSR ruling should mean the 1-click patent is overturned on grounds of obviousness.

Off-topic addendum
It is part of engineering culture to believe in the myth of the better mousetrap while deprecating marketing. Thus I was pleasantly surprised to find this quote in the KSR opinion:

. . . commercial success evidence is highly suspect, because “market-related factors, as opposed to sheer technological superiority” are theoretically and empirically much more likely to determine commercial success.

If you want your project to be a commercial success, get the best marketing you can. See The engineer’s guide to market success for some pointers.

Comments welcome, as always. Should software be patentable at all? Or should it be treated as a trade secret by developers?

{ 2 comments… read them below or add one }

Richard September 9, 2007 at 5:23 am

Robin,
This is about slowing down competition, using the cover of a very sick US patent system as a weapon. It seems that all of the related patents should never have been issued.

As per some comments on Dave’s blog, “the idea of suing for copy-on-write is like suing for using hash tables” and it is well established that “you cannot patent an algorithms” and if you go back far enough, “everyone should have to pay IBM “.

Also, we have the issue of ‘prior knowledge’, documented by Daniel Phillips …on http://uwsg.iu.edu/hypermail/linux/kernel/0010.0/0343.html

There is a crisis looming at NetApp, as supposedly their ‘defensively held’ trivia patents suddenly become “offensive”.

I believe that given Sun current positioning, the law suit is directed more at the future of open system storage, enabled by ZFS and running on ‘commodity’ hardware, very much like NetApp controllers.

ZFS is being reworked to run under FreeBSD and (I guess) it could become an option under Linux, should Sun put under GPL. With this in place, it will not be too long when various hardware vendors & system integrators are be able to offer system-level functionality similar to Sun’s “thumper”…. its not difficult.

Longer-term, from the standpoint of profit margins, OSS movement is a much more serious competition than SUN.

Sun should put ZFS under GPL and make the whole issue a more legitimate OSS case, which… as many believe, can not be won on the basis of ‘prior knowledge’.

Travis Roberts September 30, 2007 at 9:09 pm

even more ironic is that Dave Hitz himselve, describes in this blog entry, http://blogs.netapp.com/dave/2007/09/vmwares-founder.html the “inspirations” aka prior work product which “inspired” his WAFL development. Ironically netapp management has chose to interpret their founder’s use prior work product as basis for WAFL development (some of which they have patents for) as “inspiration” but interpret Sun’s use and continued development of open source ZFS as patent infringement.

Does this look a bit hypocritical to anyone besides me?

Leave a Comment

Previous post:

Next post: