Friday 28 December 2012

How to Read a Software Patent

[The article below originally appeared in another publication -- I think I wrote it about 10 years ago. I've made some updates and decided to republish here. I've omitted the original introductory texts that explained the differences between patents, copyrights and trade secret protection -- I think that's pretty well understood these days.]


Puppies help when dealing with patents.
This post describes what I've learned during "patent reviews" -- times when I've been approached to advise on the likelihood that a system we're developing infringes on a specific patent, or to assist with a patent search to look for "potentially infringed" patents.

The monopoly power conferred by a patent, the embarrassingly low standards of novelty applied by the US Patent Office, and the enormous cost of defending litigation has made patents an extremely attractive business tactic to harass or shut down competitors. It used to be the case that I would advise development teams to undertake a patent search to assess the need to "invent around" potential patent complications. However this is no longer good advice: 
  • Many patents in the field are maddeningly obvious. It soon becomes clear to anyone researching a particular technology for applicable patents that far too many software patents offer nothing more than an obvious, incremental adaptation of well known software techniques or algorithms. The worst of them do not even do that.
  • Software patents are poorly written, at least from a technical perspective. It is difficult to understand the technical architecture that the patent is trying to describe (in fact, I personally think this is a deliberate obfuscation strategy).
  • Patents take effect from the date of filing – however they do not become public until after the patent is granted. For example, US Patent 5,960,411 (Amazon’s infamous “One-Click ordering” patent) was filed in September 1997, but was not made public until two years later. This poses a problem because, at the time of an architectural review, a designer cannot be confident that they have covered all applicable patents. As the volume of patent applications increases the lag time between initial filing and final public notification is stretching to many years.
So it's doubtful whether a patent search is an economically sensible choice. Consult your IP lawyers of course. What is more likely to happen now is a business owner becoming aware of an alleged or potential infringement and asking for a opinion on likely exposure. 

Reviewing a Patent

Keep a puppy handy in case of stress
If you've been given a list of patents to review you will need to quickly scan them for relevance. Be careful here. You are sorting the patents into two lists: 
  1. The list of "obviously completely irrelevant patents that have no overlap with what we're doing."
  2. Or the "pending review" list.
See what we did there? There is only "irrelevant / not infringed" and "considering". You will never, never, write that you think "we might be infringing this one." And you will also put the words "legal in confidence" on your emails. But your lawyer will tell you all this I'm sure. :)

You'll notice that all patents have certain standard sections. Care must be taken not to rely on either the patent title, nor the patent abstract. These can be misleading and are more or less irrelevant to a question of infringement. 

As an example, Amazon’s US patent 5,960,411 has as its title:
Method and system for placing a purchase order via a communications network
So yes, the title indicates a patent that potentially covers every system of trade in the world. From the point of view of clarity, the abstract is little better:
A method and system for placing an order to purchase an item via the Internet. The order is placed by a purchaser at a client system and received by a server system. The server system receives purchaser information including identification of the purchaser, payment information, and shipment information from the client system. The server system then assigns a client identifier to the client system and associates the assigned client identifier with the received purhcaser information. The server system sends to the client system the assigned client identifier and an HTML document identifying the item and including an order button. The client system receives and stores the assigned client identifier and receives and displayes the HTML document. In response to the selection of the order button, the client system sends to the server system a request to purchase the identified item. The server system receives the request and combines the purchaser information associated with the client identifier of the client system to generate an order to purchase the item in accordance with the billing and shipment information whereby the purchaser effects the ordering of the product by selection of the order button.
I still get headaches when I read that.

What matters in a patent application is the section titled Claims. The abstract and the title have little legal significance beyond shedding light on the claims. Note that in this case (and many others), neither the title nor the abstract accurately capture the essential nature of’s patent: that the ordering process can be completed in one step (hence “One-click ordering”). 

Bear in mind when reading a patent abstract that it was not necessarily drafted by someone with any technical knowledge. Or, for that matter, actual writing skills.

Analysing the Claims

This is a Four-Puppy-Job
Having discarded the obviously irrelevant patents you will next need to work with the patent lawyer to analyse the remainder a little more closely. The outcome of your analysis will be either:
  1. The patent is not relevant because your software system can be distinguished from what the patent claims for its own. For example "our system always requires two clicks to purchase so this Amazon one-click patent isn't relevant."
  2. The patent might be relevant, but some changes can be made to make your system obviously quite different (for example, by adding that second click).
  3. The patent might be relevant. There's no easy way to make it clear that it's not. You might feel that the patent is so obvious that it should never have been granted in the first place. You'll probably be right. It won't matter though -- you'll ultimately decide what to do based on business tactics.
To arrive at the conclusions above you will be concentrating on the section titled Claims. The other sections (for example "Description") can be important (primarily as an aid to a court in interpreting the claims) -- but focus primarily on the claims.

The claims usually take the form of one or more "base" claims which cover the inventive steps. These are then supplemented by zero or more additional claims, which "inherit" from the base claim and then aim to extend and target the claims. Usually, the addition is an application of the base claim using certain specified technologies.

  1. A method of placing an order for an item comprising:
    • under control of a client system
    • displaying information identifying the item; and
    • in response to only a single action being performed, sending a request to order the item along with an identifier of a purchaser of the item to a server system;
    • under control of a single-action ordering component of the server system;
    • receiving the request;
    • retrieving additional information previously stored for the purchaser identified by the identifier in the received request; and
    • generating an order to purchase the requested item for the purchaser identified by the identifier in the received request using the retrieved additional information; and
    • fulfilling the generated order to complete purchase of the item
    • whereby the item is ordered without using a shopping cart ordering model.
  2. The method of claim 1 wherein displaying of information includes displaying information indicating the single action.
  3. The method of claim 1 wherein the single action is clicking a button.
Claims 4 –5 are also a “method of claim 1”. Claim 6 is a base claim that attempts to restate claim 1 in a wider, more generic manner and incorporates a “shopping cart ordering component.” Claims 9 and 11 are also base claims, with claims 12 to 26 being derivative of claim 11 – the widest, and most generic claim. (Note: After being granted in September 1999 there was a period of over 10 years during which the patent was challenged and re-examined.)

For the treatment of  extra-patent stress, add one kitty.
This is typical of most patent applications. If a court rules, for example, that claim 11 contains no inventive step, then claims 1, 6 or 9 (or their derivatives) may still be used to pursue an infringer. The patent application will go for an intellectual property “land grab”, attempting to have as wide and as generic an application as possible, to make it more difficult to invent around the patent or distinguish any of the claims.

The language itself will also be a barrier to understanding. The Amazon patent example I'm using is comparatively lucid. Others appear to be deliberately designed to be impossible to understand.

When conducting your analysis you will naturally focus on the base claims, since if you can distinguish your invention from each of these then the other, derivative claims will be automatically disqualified. It’s also common for each of the base claims to have some common element – get around the common element and you can get around the entire patent.

In the example used in this tutorial,’s common element is to respond to an order with only a single action being performed. This is the inventive step repeated in claims 6, 9 and 11. Putting to one side the argument about wether or not this is actually inventive you can see that by simply adding a second required action you can distinguish (that is, invent around) the patent. This is exactly what Barnes and Noble did when they were sued by in 1999.


I'm sorry for that past advice.
The original version of this essay concluded thus:
Irrespective of the debate over the legal merits of software patents, they are a reality that must be dealt with in the design and construction of any software system. This is particularly true in e-commerce, a highly valued area where competitors seek to raise the barriers to entry through prolific patenting.
A patent search should be conducted early in the design phases, and conducted a second time prior to launch to catch any additional patents and to double check architectural decisions. For the patent analysis to be conducted properly, it is important that it is conducted by a patent lawyer or intellectual property specialist.
I have to say now that this is rubbish. The volume of patents, the excessive lag between filing dates and public disclosure, and the uninspiring performance of patent examiners the world over have simply made this approach uneconomical.

In the meantime, if you're ever asked to provide an opinion on a patent, cut to the chase and look at the claims. Focus on the root claims first to find the cheapest argument out of there. If actual cease-and-desist letters are being sent though, call in the lawyers.

Sunday 23 December 2012

A Brief History of Trademarks

Of all the regulatory systems that get lumbered with the umbrella of “intellectual property”,  Trademarks is the one with the oldest history.

Popular accounts date trademarks back to the troglodytes. Troglodytes loved their trademarks apparently. Cave paintings demonstrate the importance of leaving one’s mark on things, in much the same way as Coca-Cola likes to leave its mark on bottles of fizzy sugar water.

You can probably tell that I’m a bit skeptical.

Personally, I think the cave paintings show a deeper need. An expression of identity, a questioning of “who am I?” An attempt to leave a mark upon the Earth to say “I was here” that will last longer than the brief flash of life would otherwise have us stay.

Other authors see a desire to establish a brand presence. Well oh-kay…

Trademark history is also lacking a character. Copyright has its characters, as does Patent Law. However trademark history seems to lack anyone of any real character or interest.

Well, except perhaps Thomas Jefferson.

Pre-Jefferson Trademarks

Bass Brewery's logo 1875
Trademarks are old (I just don't think they're Palaeolithic old). Evidence of their use exists in both medieval Europe and Ancient Rome[1]. The first registered UK trademark is that of the Bass Brewery, a simple red triangle registered in 1876. The first US registered mark was for Samson Rope in 1884. The logo is still used today.

Earlier examples also exist, for example in Ancient Egypt & China. The marks are generally taken to be an indication of origin, to both take credit for the quality of the work but also perhaps to indicate who to blame if there was a defect.

Gary Richardson has argued that subtler systems of product differentiation may have served a larger purpose than simply indicating the origin of the product. To quote the abstract from the relevant paper:
"In medieval Europe, manufacturers sold durable goods to anonymous consumers in distant markets [...] by making products with conspicuous characteristics. Examples of these unique, observable traits included cloth of distinctive colors, fabric with unmistakable weaves, and pewter that resonated at a particular pitch. These attributes identified merchandise because consumers could observe them readily, but counterfeiters could copy them only at great cost, if at all. Conspicuous characteristics fulfilled many of the functions that patents, trademarks, and brand names do today. The words that referred to products with conspicuous characteristics served as brand names in the Middle Ages." [See [1], below. Emphasis added.]
Trademark law really got cracking however after the industrial revolution (and that's turning out to be an interesting inflection point in the history of IP but I digress...). And that's where Thomas Jefferson comes in.

Thomas Jefferson

Thomas Jefferson.
In 1791, a group of Boston sailcloth makers petitioned Congress to be allowed to register their marks. The matter was referred to Thomas Jefferson[2], then Secretary of State, who decided:
That it would, in his opinion, contribute to fidelity in the execution of manufacturing, to secure every manufactory, an exclusive right to some mark on its ware, proper to itself.
That it will, therefore, be reasonable for the general government to provide in this behalf by law for those cases of manufacture generally, and those only which relate to commerce with foreign nations, and among the several States, and with Indian tribes.
The sailcloth makers were concerned that they would be undercut by makers of inferior quality sails, who would misuse the Boston marks to mislead customers. Customers would buy the cheaper sailcloths and develop a poor view of the Boston makers whose marks had been used on them.

Until the government legislated for the exclusive use of registered marks, the most common remedy from the law would have been to sue for damages. This however was probably ineffective -- counterfeiters continued at will. A letter printed in the Columbian Sentinel was fulsome in its praise of Jefferson's decision:
[I]t is of the greatest importance to the rising prosperity...[and greatness] of this country, that the manufacturer should be secured in the benefit and profit of his ingenuity, labour and industry, being an carry on, persevere in, and bring to the greatest possible perfection the various goods and articles [they make].
There is no greater check to this laudable spirit of enterprise...than that of impostors fraudulently counterfeiting of marks...and selling bad and spurious articles for good, real, and genuine [goods]. It effectively cools the ambition of selling...and is highly prejudicial to the good repute of our manufacturers in foreign parts, [lessening our countries commerce].[4]
TJ took a much more positive position towards trademark protection than he did for copyright or patents. The restrictions on liberties that were threatened by those systems were still there -- however this time there were no serious arguments that such liberties had to be guarded. For copyright and patents, the argument was that the restraint on trade was outweighed by the need for limited monopolies in order to create incentives to work. For trademarks, no such trade-off exists. That's because trademarks were seen as incentives for everyone to work. There was nothing to stop honest manufacturers creating their own marks and practicing the same trade. The only prohibition was that they could not represent their goods as originating from somewhere else.

Justification for Trademarks

So a system of registered trademarks was good because:

  • It encouraged pride in workmanship, since it linked manufactured goods with a manufacturer's reputation.
  • It protected the national economy from having the reputation of its industry undermined by poorly made knock-offs.
  • It protected the public consumers from being deceived by counterfeiters. 
Recall that, in the development of patent law, the right and ability to work was part of God's Law. Therefore, patents had to be limited so as to not interfere with the will of God. Trademarks also supported this concept of God's Law, since protecting ones reputation (and isn't reputation everything?) was critical to safeguard the right and ability to work. Trademarks were not seen as potentially unjust encroachments on liberty... not even by the likes of TJ himself.

These developments in trademark law were post the industrial revolution. The arguments justifying the changes were entirely instrumentalist and economic in nature. No metaphor used at all.

Finally, I think I might wrap up this series with a quick look at Designs, Plant Breeder Rights and other sui generis protection systems.

Wednesday 12 December 2012

It's Beginning to Look a Lot Like Christmas...

Image of subclavian and axillary veins in arm
Veins of the Upper Arm
Yes, it's that time of year when I like to go into hospital and have a rib removed.

For the curious:

  • Paget-Schroetter Disease is when a blood clot (thrombus) forms in the arm due to the first rib and clavicle and the muscles in the area squeezing the vein that runs out of the arm and causing it to "kink" (a bit like a hose). This causes the blood flow to slow which leads to a clot forming, which causes the blood flow to slow...
  • The rib-clavicle-muscle issue is known as Thoracic Outlet Syndrome (TOS) and is a compression of the outlet pathway between the arm and chest. There are multiple causes for this. Mine is the "poorly designed rib & clavicle" kind.
  • Isaac Hanson had a pulmonary embolism as a result of TOS. 
When it first happened to me in October 2010 I was in ICU for 10 days while the clot was disolved using thrombolytics. It's not so much the clot that's dangerous at that stage as the drugs used to treat it. You're in ICU so that they can monitor it. After a stent was inserted in the subclavian vein (veins have a "shape memory") I had my first rib "resected" -- cut up into little pieces and removed. The TV series RPA featured the procedure (by the same surgeon treating me) in that aired in September 2010 (a few weeks before I presented at RPA with the same problem). Watch it, it's fascinating. :)

Monday 10 December 2012

A Brief History of Patents

You have to feel a little sad for Filippo. He was an amazing artistic, engineering and architectural genius. He was revered in his own lifetime. He had convinced his government that he had an amazing and revolutionary invention that needed an extraordinary protection if he were to go ahead and build it...

But mere hours after its launch his creation, his pride, and a sizeable part of his fortune lay at the bottom of the Arno river.


Portrait of Filippo Brunelleschi
Filippo Brunelleschi was a dude. I think he'd be quite popular today. He was a talented and gifted architect, artist, and engineer. He was commissioned to build the dome for the incomplete Santa Maria del Fiore, the Florentine cathedral, after he won a competition which involved standing an egg on its end on top of a piece of marble.

He won the competition by flattening one end of the egg by breaking it slightly. It sat quite easily after that. When the other competitors complained that they could have done the same thing, he reportedly laughed that if they could have done that then they would have built the dome already.

See? Dude.

To complete the dome Filippo had to invent new hoist systems to lift marble blocks to the height required. He also had to rediscover how to build domes that large, since no-one had been able to do so since the Romans. Finally, he had to ship a mind-boggling shitload[1] of marble into Florence.

To do this, he invented a new kind of ship: Il Badalone (The Monster). Frankly just naming the thing ought to have cemented his reputation as a marketing genius as well. However, Filippo was concerned that others would simply copy his ideas, so before building it he petitioned the State of Florence for a three year monopoly on the concept, after which he would build the ship and hoists and so reveal his secrets.[2]

Filippo was so influential that he didn't even need to reveal the design beforehand. The authorities trusted that he was onto a winner -- after all, he had form. The document is extraordinary by modern standards. It basically freezes the state of the art of ship building for three years, saying that only Filippo Brunelleschi may sail ships of a new design. Anyone else sailing any kind of ship on any kind of water that was not already in existence before Filippo's design would have their vessel burned. Only Filippo himself could design, build and sail new types of craft.[3]

Drawing of ship suspected to be Il Badalone
Il Badalone
When the ship finally set off on its maiden voyage in 1428 it sank, taking 50 tonnes of marble to the bottom of the Arpo river. It had sailed only 40 kms.

From everyone's perspective, this was surely not a great bargain. The state imposed a monopoly cost on its citizens, in exchange for which, it gained knowledge on how to build a vessel that could sink very efficiently. It won't be the last time however that the patent quid pro quo turns out to be a lousy deal.

The monopoly on new ship designs granted by Florence was an informal exercise of the state's prerogative. The Republic of Venice codified the principles under which such monopolies would be granted in 1474. This time, it was expected that inventors would reveal their inventions once perfected, and then enjoy monopoly rights over it for ten years.[4]

Meanwhile, in England...

The Royal Prerogative to Do as Thou Will

A prerogative is a special right or privilege. The word seems a little out of place when talking about monarchies (after all, isn't that what monarchies are all about? I mean, is there another kind of royal privilege?). Nevertheless, the history of democracy is basically about the gradual restrictions imposed by parliaments on Royal Prerogative.

Typically the granting of this right was done in writing, by way of a royal letter. The letter was addressed to all the subjects of the realm and was therefore an open letter. "Patent" (as most med-students or people with stents[5] would know) means "open". Hence there were "letters patent" -- an open letter to the public at large that granted a special right or privilege to another.

Portrait of King James the first
King James I
In mediaeval England, the monarch could grant citizens a monopoly over pretty much anything. Monopolies tend to be enormously profitable and so were very helpful in raising taxes while deflecting blame. For example, James I granted exclusive rights to "export calfskins; ...import cod and ling; [and] make farthing tokens of copper."[6] The monopoly holder could charge very high prices because of the monopoly, but would also be expected to pay a large tax on the revenue. Even commodities like salt were restricted in this way for a time. The monopolist got the blame -- the crown got the revenue.

Prerogative based monopolies were very powerful. Royal prerogative is subject to both the Magna Carta and the common law courts. The courts eventually got their chance to set out the requirements for a valid monopoly in the Case of the Monopolies.

Portrait of Queen Elizabeth the first
Queen Elizabeth I
Queen Elizabeth I had granted an exclusive license to make and sell playing cards to one Edward Darcy. The case suggests this was out of concern that the lower classes ("servants and apprentices and poor Artificers") were spending too much time playing cards and not enough time learning a trade. A monopoly on printing and distributing playing cards was issued to one Ralph Bowes for twelve years, after which it was then granted to Darcy. The grant allowed Bowes and Darcy to restrict the distribution of cards, and also involved the payment of fees to the Crown.

Thomas Allen must not have seen the memo. He was sued by Darcy for selling playing cards without his permission[7]. Ultimately Allen prevailed -- the Queen's Bench found that Darcy's monopoly was invalid. It was not a proper exercise of the royal prerogative, because:
  • Every citizen had a right to a lawful trade. A monopoly restricted this right, and was therefore not in the interests of the commonwealth. The prerogative right to issue monopolies could not be used if it harmed the citizens. But of course that's exactly what a monopoly does: it raises prices; restricts livelihoods; and impoverishes people who were trading prior to the monopoly existing.
  • It was argued that it was part of God's Law that men must labour to support their families. Therefore to restrict this ability was to go against God. This was consistent with Church teaching "which had condemned monopolies in the strongest possible terms."[8] A monarch could not use the royal prerogative to injure their subjects, for they themselves must obey God.
There was one exception allowed for by Allen's counsel. If a person invented something new, or was offering to bring foreign inventions to England and to teach their proper use to others, then a monopoly right could be lawfully given. This was because the whole realm benefited from new trades and  technologies. There was a corollary: once learned, a monarch could not renew the monopoly, because the benefit was no longer available (since the knowledge was now available to everyone).

The abuses of prerogative was one of the focal points of the struggle between the crown and parliament. This particular battle was finally resolved by the Statute of Monopolies in 1624.

Statute of Monopolies

Sir Edward Coke
The arguments against the establishment of prerogative monopolies, including the exception for inventions, was codified in 1624 in the Statute of Monopolies. The statute first revoked all monopoly privileges then in existence, since they were not in the public interest. Only one exception was allowed: patents for inventions for new methods of manufacture. However even these had an expiry date, being 14 years after issue.

Parliament rejected a "natural law" justification.  These patents for inventions were justified purely on economic incentives. Even in America and France, where there was no history of royal abuses of the privilege system, the justifications for monopolies on inventions was based on instrumentalist arguments and economic incentives. This was despite the fact that in both countries, the concept of natural rights, liberties whose existence where held to be self-evident, had no such restrictions.

The statute was quite short and broad -- it was up to the courts to interpret its application. Sir Edward Coke, writing in the Institutes of the Lawes of England, summarised the requirements for a valid patent:
[N]ew manufacture must have seven properties. First, it must be for twenty-one years or under. Secondly, it must be granted to the first and true inventor. Thirdly, it must be of such manufactures, which any other at the making of such letters patent did not use ... Fourthly, the privilege must not be contrary to law ... Fifthly, nor mischievous to the state, by raising the prices of commodities at home. In every such new manufacture that deserves a privilege, there must be urgens necessitas et evidens utilitas. Sixthly, nor to the hurt of trade ... Seventhly, nor generally inconvenient.
In modern terms, we might say that a valid patent:
  1. Is a temporary privilege, never a perpetual right.
  2. Must be granted to the first inventor.
  3. Must be new, and not cover the existing state of the art.
  4. Must not be for an invention that would be otherwise against the law.
  5. Nor can it be something that harms the state (for example, by raising the prices of commodities).
  6. The invention must be for something useful or the patent urgent and necessary.
  7. The patent must not be generally inconvenient.

Modern Patent Justifications

Graph showing steady increase in software patents from 1985 to 2000
The fact that the formula for a valid patent has remained virtually unchanged for almost 400 years is interesting in itself. There does seem to have been a subtle shift though in the arguments for the existence of patents. Whereas the early English cases speak of increasing the "knowledge capital" for the public good, modern justifications emphasise the need to ensure adequate returns on investments in research and development. 

Also in the early history of patents, the use of physical property as a metaphor to guide the law does not exist. The issuing of a patent is given in almost reluctant terms: that the state would really rather not issue these temporary monopolies, but will grudgingly do so to improve the knowledge of the nation. To do this, inventors may be granted temporary monopolies, as long as they disclose an invention that is new, useful, and does no harm.

Does that sound like the system we have today?

A thought occurred to me while reading Drahos and writing this post: that metaphor is powerful precisely because it constrains. It creates a framework which contains the subject being described. If the language of IP was still about privileges to promote the creation and sharing of new knowledge for the public benefit, would patent maximilists have been as successful in expanding and extending patent law?[9]

I've spent too long on this post already! The next thing to look at will be trademarks.