MIME creator: Developers face patent trap

Patent infringement lawsuits have become a very real concern for independent developers and big companies alike, a situation highlighted by the growing number of high-profile courtroom battles over intellectual property.

At the moment, for example, Oracle has sued Google over Android's implementation of the Dalvik virtual machine, and Apple is fighting a number of smartphone competitors in several countries over patents related to the iPhone.

The concern is that broad software patents, and the risk of litigation these bring, could stifle innovation by developers. With this in mind, ZDNet UK sat down with Nathaniel Borenstein, one of the inventors of the MIME email protocol, for his take on the matter. Borenstein is chief scientist at cloud email management company Mimecast, and his background includes eight years as a distinguished engineer at IBM.

Q: I understand Mimecast is actively working on technology to allow it to build email analytics capabilities for organisations?
A: We're at the very beginning stages of it. [In October] we filed a patent application, so we're at least working on it that much. I wouldn't say we've done a lot of coding yet or a lot of decisions about what to deploy. We're the classic small-to-middling company growing too fast to do too many new things. It's not going to happen right away — we're still fleshing out the API potential.

What did the patent application cover?
It's basically a business model patent, filed in the US, on cloud-enabled applications. It's very broad. I have a pretty deep ambivalence in that I think the patent system is totally screwed up and broken. All sorts of overly broad and inappropriate patents are granted, and this could be another one of them.

Well, this sounds like a pretty broad one itself...
If we didn't file it, somebody else might, so I'd much rather we file it than one of our competitors. I'll be the first person up there to explain why the patent system shouldn't allow things like this, but it does.

Why do you think these kinds of general patents shouldn't be allowed?
The patent system was designed to foster certain kinds of innovation by creating a protected time for limited commercial monopoly. The system has evolved where that time has become inappropriately long, particularly in our industry.

So you think it should be more like the pharmaceutical industry, where the time frames are shorter?
Well, in the US they still have a longer period, but it's more appropriate for them. First of all, they have to spend billions of dollars to create the drug. Second, it's really easy to see if someone is violating their patent: look for the molecule — it's that simple.

The patent system is totally screwed up and broken. All sorts of overly broad and inappropriate patents are granted.

With software, it's all very vague and general. You can say, "Well, here's a patent that uses the same words as you use in your product, so you're infringing on it." They don't actually say that, but that's essentially what they are saying. That leads to a horrendous waste of resources like the Apple-Google battle, which is actually Apple versus Google, Samsung [and] HTC, at least.

It's a terrible waste of resources. It's not appropriate to give 17 years' protection for a concept as broad as the iPhone. Do you really want them to have a monopoly on smartphones for 17 years? I don't think that's the kind of thing people had in mind with the patent system. If you can't make money on smartphones in three years, then you probably aren't doing it right.

The business model patents are even worse. My favourite is that there is a patent on a business process in which you patent things about business processes and then sue people that infringe them.

That's nuts! And that was granted?
No kidding! That was granted!

The people that get that, what do they do with it? It's a very tricky business. You don't even try to enforce it, per se. Here's the thing: once you get into a patent battle, the winner of the battle at a first pass is normally determined by who has more money for lawyers. It's only when it's closely matched, like with Apple and Google, that it gets to a higher level.

If you have that patent, the first thing you do is...

...make sure it is assigned to a non-practising entity (NPE). The point of an NPE is that they don't do anything but sue over patents, which means they can't be sued for much of anything else. You can't sue them for violating a patent, except maybe that one I just described. Then that NPE looks for targets.

So you've got a patent on a more-or-less bogus thing, what do you do? Do you sue IBM because they have the most money? No, because they also have the most lawyers, and the worst thing that could happen is that they could actually have a lawsuit go through and invalidate the patent. You give those guys a free pass and instead look for someone small enough that they'll be frightened, and you sue them. Then you offer them a settlement that's much more attractive.

This happened at Mimecast last year. We got sued by people who basically had a patent on sorting by keywords. I looked at it and said, "I was doing this in the '80s, there's no way this would hold up in court." We spoke to our patent lawyers who said, "You're probably right — you're the expert, you could probably defeat it in a validity battle." But sometimes it's a situation where 'winning' and fighting the patent claim can cost, say, 10 times more than 'losing'.

Doesn't that suggest Mimecast is a soft target for patent battles? Isn't there a point where a business might want to say, "You know what, we're not going to roll over"?
That's a matter of principle. There's another principle, which is the responsibility to the shareholders. Index Ventures invested tens of millions of dollars in Mimecast, and frankly we asked their opinion, because it's their money. I'm not sure of the ethics of spending it on that, either — there's just no good answer there.

There are proposals for a Europe-wide patent system. Is the US looking at reforming the patent system there?
The US system is the worst, both because of the system itself and because we're such a litigious society. The US just passed a patent reform which didn't solve anything that I consider to be the major problems.

The US just passed a patent reform which didn't solve anything that I consider to be the major problems.

To give an idea of what passes for reform, one of the provisions is when you file a patent there's a filing fee, then you wait a couple of years. You can now pay a bigger filing fee and get bumped to the front of the line; that's a reform. That struck me as one of the most illuminating parts of the reform.

The problem is that the people who lobby Congress are the big companies who are the major patent holders and they have, almost against their will in some places, twisted into becoming advocates for the status quo.

IBM is a great example. It has been the largest patent receiver in the world for something like 20 years running. Their strategy was initially conceived, and is still conceived, as primarily defensive — they do this so that if somebody sues them, they can sue them back. But some of the patents are so valuable they make licensing money from them. Last I heard, those patent licensing revenues were over $1bn [£625,000] per year — that's incidental, it's not their primary purpose.

It seems the only people who win are the lawyers.
No, the only people who are guaranteed to win are the lawyers — sometimes other people do win. Nowadays, they even try to obfuscate what the patent is about, because the more obfuscated it is, the more they can claim on it. In the first three decades of the computer industry, software was not patentable.

So what can be done about this? Does it not deter software innovation from independent developers?
Well, for example, I have a non-Mimecast idea that I'm sure is patentable. I've been given permission to work on it outside the company, if I want to spend money on lawyers, which I haven't decided. My biggest concern about not doing it is that someone else will. I'd like to do it and collect the royalties, or do it and put it in the public domain — but I can't afford to spend a lot of money to put something in the public domain either. So that's the situation I'm struggling with right now.