Opinion piece - Innovation

What is innovation?

The dictionary definition is fairly bland. Dictionary.com defines innovation as "Something new or different introduced: numerous innovations in the high-school curriculum. The act of innovating; introduction of new things or methods." I have no idea which dictionary OS X uses but it defines innovation simply as "a new method, idea, product, etc. : technological innovations designed to save energy."

The OS X dictionary is, I think, slightly more revealing and in line with what people normally understand by innovation. From the Dictionary.com definition, innovation could simply be regarded as newness for newness sake. The example given in the OS X definition – new technological innovations designed to save energy – has an implied additional context of newness with a purpose, of newness with a tangible benefit. This also implies some kind of standard to be met, a bar to be cleared if you like, before something can be thought of as innovative.

There are a lot of differing opinions about where that bar should be set. On the one hand you have the die hard technology geek. Innovation to them is something big and brand new, something that has never been achieved before. A fusion reactor, a flying car, a quantum computer – that kind of thing. Internet forums inhabited by the die hard geeks tend to be full of posts along the line of 'yeah, not bad but wake me up when it does this'. Or this ever popular sequence:

'Company A has released this incredible new technology – wow!'

'Meh – that's just a tweak on what company B was doing years ago – company A is just copying them.'

'Pfff – Company B got all their best ideas from Company C. Besides, real engineers know that Company D did it first and anyway product X by company E is what you really should be using.'

At the other end of the spectrum you get the popular press and marketing hype where every incremental improvement is trumpeted as 'Innovation'. Using the dictionary definition, I suppose it is but that strikes me as sticking to the letter rather than the spirit of said definition. Naturally the die hard geek dismisses this level of innovation as 'marketing' – a catch all dismissal of any product that manages to be far more popular than a competing product, even though it is technologically 'inferior'.

(Yes, this post is full of sweeping generalisations. That's why it's marked as an opinion piece.)

Personally I think it would be a much saner world if we could avoid either of those extremes. It would certainly save a great deal of time, energy, paper, marketing budgets and recycled arguments. Most innovation consists of incremental improvements on a concept, or the repurposing of old concepts, rather than brand new concepts. To quote Ambrose Bierce - “There is nothing new under the sun but there are lots of old things we don't know.”

Loudly proclaiming the 'innovation' of each and every advance or minor variation on an old theme isn't helpful and simply degrades the word to the point where the bigger advances or genuinely useful reworking of old themes get lost in the noise. On the other hand, dismissing everything that isn't absolutely brand new as not being innovative does a similar disservice to those same advances or reworkings. To my mind, taking an existing but little used product, working out why it's not used and then designing a similar product to overcome those barriers to use, most certainly qualifies as innovation. It might not be a brand new product, it might not be exciting to the purists or technological connoisseurs but it can certainly be innovative.

Another day, another troll

A number of news sites are reporting that Smartphone Technologies (a subsidiary of Acacia Research) have filed a lawsuit against Amazon, claiming infringement of four of their patents. Most of the sites seem to link back to this article at paidContent.org. Details of the actual infringed patents are somewhat sketchy, although US 6,956,562 is reported as one of them.

Acacia is widely regarded as a patent troll, i.e. a company that does not sell any products or services itself and buys patents solely to monetise them through licensing or litigation. The lawsuit was launched in the Texas Eastern District Court - so no surprise there either.

Cue the normal spate of predictable anti-patent commentary. However the paidContent article itself seems a little unclear about what is being infringed and why. From the article:

U.S. Patent No. 6,956,562, for instance, seems to describe the act of tapping an icon in order to instruct the device to perform an action:

According to the method, a graphical feature having a surface area is displayed on a touch-sensitive screen. ..To control software executing on the processor, a user-supplied writing on the surface area is received and the software is controlled responsive to the writing.

To me, this looks more like a method for controlling a smartphone, tablet, PDA etc by handwriting recognition. Not quite the same as tapping an icon.

So lets have a look at the actual patent. Published 18 October 2005 so fairly old and has a priority date of 16 May 2000, which isn't ridiculous either. This doesn't appear to be an ersatz submarine patent with umpteen continuations or continuations in part. Lets check the claims - 66 of them with only two independent claims. Again, not too ridiculous - I've seen far worse. Claim 1 reads:

A method for software control, comprising:
displaying a graphic representing a set of one or more computer functions on a portion of a touch-sensitive screen, wherein the touch-sensitive screen is coupled to at least one processor to detect and interpret contact with the screen; 
detecting an object making a first sequence of one or more contacts that form a first drawing on the portion of the screen; 
in response to detecting the object making the first sequence of one or more contacts that form the first drawing: 
matching the first sequence to a particular action in a set of actions, and performing the particular action; 
detecting an object making a second sequence of one or more contacts to form a second drawing on the portion of the screen; 
in response to detecting the object making the second sequence of one or more contacts to form the second drawing: 
matching the second sequence to a second action in a set of actions related 
to said one or more computer functions, and 
performing the second action; 
wherein the visual appearance of the graphic is the same when the first sequence of contacts is commenced and when the second sequence of contacts is commenced. 


Claim 47 is a claim to a computer controlled by the method of claim 1.


OK, displaying a graphic representing one or more computer functions on a portion of a touch sensitive screen. Yep, that sounds very much like an icon. Making a sequence of one or more contacts to form a first drawing - I'm not sure that this does encompass a simple button tap. You could argue that the claim encompasses a single contact - aka a tap but can a tap be reasonably considered to be a drawing? In my opinion no. However, writing a letter on the screen could be construed as a drawing. As for the one or more contacts - consider the difference between the letters 'o' and 'i'. The first can be drawn with a single contact, the second requires two separate contacts to make the line and the dot.

Furthermore, claim 1 is quite limited (presumably to avoid prior art) in that it stipulates a chain of two commands (both input via drawings as discussed above) and that the the appearance of the icon (or other part of the GUI) remains the same between both commands.

All in all, quite an odd little patent and certainly not a simplistic 'patent for tapping an icon on a smartphone'. Actually this reminds me more of Swype - where words are entered in e.g a text message by performing a gesture across a touchscreen keyboard. Entering a word into a text message could also be regarded as an 'action' in the context of the claim....

Opinion piece - copyright duration.

I'm a couple of weeks late with this one but as reported by numerous news sources, the copyright term for performers and sound recordings has been extended by 20 years to 70 years. By way of background, in Europe, the copyright in artistic or literary works (which are fairly broadly specified in Article 2 of the Berne convention) lasts for 70 years after the death of the author. The copyright in sound recordings and broadcast performances (whether by radio or television) lasts for 50 years. Copyright durations in the UK are in line with European copyrights.

Personally, I think that the extension of copyright in sound recordings is disappointing – or rather the almost complete lack of debate on the matter is disappointing. Two government reviews of intellectual property law (Gowers and Hargreaves) both concluded that there was little or no economic case to be made for extending copyright durations. Despite this, the extension was waved through by the government without so much as a murmur of dissent. So much for evidence based policy making.

I'm not alone in this judging by the trenchant commentary on other blogs.

Speaking personally, I also believe that current copyright durations are far too long. Don't get me wrong – I think copyright in general is a very good thing and I certainly don't begrudge writers, musicians and performers of all stripes (artists for short) the ability to make money from their work – and in some cases to make an awful lot of money from their work. However, I disagree with the justifications put forward for the current excessive duration of copyrights.

All intellectual property rights, should be a balance between a benefit to the public and a benefit to the rights owner. In the case of copyright, artists are given a legal tool which allows them to make money by distributing copies of their work whilst giving them the right to prevent anyone else from copying that work. This is balanced by two things – firstly that the copyrighted works will eventually pass into the public domain and secondly that the possibility of making money encourages more people to create artistic works for the public to enjoy.

Nowhere in this can I see a reason that copyright should be a gravy train for an artist's descendants and for their descendants too. However this is precisely what the legislation specifies. Before the last round of extensions, copyrights lasted for 50 years after the death of the author, which was 'intended to provide protection for the author and the first two generations of his descendants'. It was then noted that people were living for longer and lo and behold – a justification for extending to 70 years after death was found.

On a similar note, one reason for the recent extension of copyright in sound recordings is (and I quote) “performers generally start their careers young and the current term of protection of 50 years applicable to fixations of performances often does not protect their performances for their entire lifetime. Therefore, some performers face an income gap at the end of their lifetime.”

Seriously? Most people face an income gap at the end of their lifetime. It's called retirement. A lot of people from all backgrounds and incomes also have the wisdom and foresight to save something for retirement. Why shouldn't performers?

Sarcasm aside, it's worth having a look at this in a bit more detail. The average retirement age (defined as the age at which state pensions become available) in the EU is just over 65. I'll say 66 for convenience. So a copyright duration of 50 years would last a performer from the age of 16 to an age at which most other workers would recognise as a retirement age. That doesn't sound so harsh to me. Extending the copyright to 70 years covers a performer from birth to well past that retirement age – which strikes me as being plenty for even the most precocious of young performers.

However, linking copyright duration to working age isn't actually a bad idea. What if that applied to all copyrights? Forget about 70 years from death – just a single fixed term of 70 years from when any artistic work was created. In practice this would cover most artists for their whole lifetime. Even for artists that start their careers at a very early age it would last for their working lives, or at least for a duration that workers in every other field would recognise as a working life. 

In many cases 70 years from creation would mean that copyright in an artistic work would still extend beyond the lifetime of the artist but not to the ridiculous extent that it presently does. For companies that invest in artists and their works, 70 years isn't an unreasonable time to recoup that investment either. Granted it would be significant reduction from what they currently enjoy but forgive me if I believe that these companies should be investing in new talent, rather than resting on the laurels of artists long since dead and gone.

The only significant losers are the artist's descendants and their descendants too. Again, forgive my lack of sympathy but to them I say welcome to everybody else's world. Enjoy whatever is left to you in your grandparent's will and then go out and earn your own money.

Minnows and Giants

Giants and Minnows

I was flicking through the BBC news site a couple of days ago and found this article. I was curious to find out what this was all about and so decided to find US patent 7,222,078 and have a look for myself.

This is certainly not a short patent, with 39 pages of drawings and a further 50 pages of description. The abstract isn't exactly a model of clarity and brevity either. So lets turn to the claims. As I posted a while ago on this blog, the claims are the most important part of a patent because they define what the patent is preventing you from doing.

Claim 1.

A system comprising: units of a commodity that can be used by respective users in different locations, a user interface, which is part of each of the units of the commodity, configured to provide a medium for two-way local interaction between one of the users and the corresponding unit of the commodity, and further configured to elicit, from a user, information about the user's perception of the commodity, a memory within each of the units of the commodity capable of storing results of the two-way local interaction, the results including elicited information about user perception of the commodity, a communication element associated with each of the units of the commodity capable of carrying results of the two-way local interaction from each of the units of the commodity to a central location, and a component capable of managing the interactions of the users in different locations and collecting the results of the interactions at the central location.

Claims 2-53 (yes seriously) all claim various extra bits and pieces added on to the core invention claimed in claim 1. Claim 54 reads:

A system comprising: units of a facsimile equipment that can be used by respective users in different locations, a user interface which is part of each of the units and is configured to trigger a two-way interaction to occur on-line between the unit of the facsimile equipment and a vendor of the facsimile equipment, the user interface being configured to generate information about use of the unit by the user, a communication element associated with each of the units capable of carrying results of the two-way local interaction from each of the units to a central location, and a component capable of managing the interactions of the users in different locations and collecting the results of the interactions at the central location.

OK this is interesting in that it claims a specific kind of system, namely a fax machine, that uses a similar system to that claimed in claim 1.

Claims 60 claims another variation on the general theme of claims 1 and 54 and claim 69 claims a method of using a similar system to that claimed in claims 1, 54 and 60.

I'm not going to pretend to do a full analysis of all 74 claims here but I can try and dissect the first claim. My first thoughts were that this has been done before and that it sounds very much like the kind of help system of the kind that's built into Microsoft Office software – Word 2003 for example. So lets break down claim 1 into its separate parts and see how well that fits.

Units of a commodity that can be used by respective users in different locations,

A software package would fit that definition and I would argue that 'units of a commodity' is general enough to include parts of that software package too.

a user interface, which is part of each of the units of the commodity, configured to provide a medium for two-way local interaction between one of the users and the corresponding unit of the commodity. 

The Word 2003 help system provides for two way interaction between a user and the software package. To be honest I'm struggling to think of a user interface that doesn't allow for this – isn't that the point of a user interface?

and further configured to elicit, from a user, information about the user's perception of the commodity, 

Well the Word help system certainly does this – a given help page usually has a 'was this useful' option at the bottom.

a memory within each of the units of the commodity capable of storing results of the two-way local interaction, the results including elicited information about user perception of the commodity, a communication element associated with each of the units of the commodity capable of carrying results of the two-way local interaction from each of the units of the commodity to a central location,

If I remember correctly, the Word 2003 help system does indeed report any response to the 'was this useful' dialog back to Microsoft aka a central location. It's debatable whether 'a memory' and a 'communication element' are limited to actual hardware or whether they would also include software equivalents. Given that the owners of this patent are trying to enforce it against a smartphone app developer, they clearly think that software equivalents are fair game, so lets go with that.

and a component capable of managing the interactions of the users in different locations and collecting the results of the interactions at the central location.

This is where my analysis gets a little more speculative, although (as we'll see later) there's a much bigger elephant in the room anyway. The Word 2003 system isn't all stored on the end user's computer – some of the help files are stored remotely and delivered over the Web. I would guess that the web based help files are also updated in response to received feedback via the 'was this useful' dialog. I would therefore argue that this web based system is a component 'capable of managing the interactions of the users in different locations'. We've already seen that the system is capable of collecting interaction results at a central location.

Therefore, at first glance, it does look like the Word 2003 help system meets all the requirements of the patent claim. This is useful for giving a picture of the kind of things covered by the patent but isn't very helpful otherwise because I'm basing the analysis on a software package released in November 2003. However, the patent we're discussing (US 7,222,078) has a priority date of August 1992. What is a priority date and why is it important?

Under certain circumstances it is possible to file a patent application that claims priority from an earlier application. That means that the new application is treated as having the same filing date as the earlier application. For example in the UK, it is possible to file a first patent application and then, within the next 12 months, to file a second application based on the first that claims priority from it. This is useful because it allows inventors to file an early patent application for an invention that may not be fully worked out, or may need substantial development funding before it will be commercially viable. They then have 12 months to carry out further development work, raise venture capital and generally decide exactly what they want to do with their invention before having to spend more money on their patent application.

However you only get one go at this. It isn't possible to chain applications together, for example by waiting another year and then filing a third application based on the second that also claims priority from the first.

The US system is somewhat more involved and as I understand it, under certain circumstances it is quite possible to chain applications together in this way. This is exactly what we see with US 7,222, 078. Although the patent itself was filed in December 2003, it claims priority from another application filed in September 1997, which in turn claims priority from another application filed in May 1994. The May '94 patent claims priority from a still earlier patent but was filed under slightly different rules (a continuation-in-part rather than a continuation). I'm not going to go into the details of the rules here but the upshot is that only part of the invention described in the May '94 patent will be entitled to that earlier priority.

In summary, US 7,222, 078 will have a priority date of at least May 1994. This is important because only publically available information from before that date can be used as prior art, ie. material that can be used to argue that the patent shouldn't have been granted. Therefore, although US 7,222, 078 was filed in December 2003, Word 2003 (which was released before that date) cannot be used as prior art because US 7,222, 078 is treated as though it were filed in May 1994 (and possibly even earlier).

So if we want to demonstrate that US 7,222, 078 shouldn't have been granted, we will ideally need to find evidence of a system that was in use before 1992 that includes all the features set out in claim 1. The good news is that since US 7,222, 078 was filed after June 1995, it only lasts for 20 years from its priority date. So at worst, it will automatically expire in 2014 and may even expire next August.

Back in 1992, the invention described in US 7,222, 078 may well have been cutting edge fax machine technology and quite deserving of a patent. It's unfortunate that the patent was drafted widely enough that 20 years on it is being enforced against inventors working in rather different areas of technology. In fact this case rather reminds me of the the BT hyperlink patent, where an old, broadly drafted patent was optimistically asserted to cover much newer technology.

The BT patent was found not to be infringed by the World Wide Web and the issue of prior art didn't need to be addressed, although some was found. It is unclear what will happen with the Lodsys patent.

Patents vs Designs

Of all the technology patent lawsuits in progress at the moment, the Apple vs Samsung case in Germany (iPad 2 vs Galaxy Tab 10.1) has fed the fires of many an internet rant. The idea that Apple could get a 'patent for a rectangle with rounded corners' has provoked widespread incredulity and has been painted as (another) sure sign of a broken patent system.

There's just one problem. The 'rectangle with rounded corners' patent is a design patent.

Although that looks like a prime example of legal hair-splitting, it's actually an important distinction. In the US, a utility patent is what most people are talking about when they talk about patents. Utility patents protect new inventions and contain a technical description of how that invention works. Design patents on the other hand protect the three-dimensional shape of an object. The aesthetics or 'industrial design' if you like.  Unfortunately (although understandably) this distinction is usually lost and design patents tend to be seen as just another utility patent. Hence all the irate commentary.

In Europe, things are a bit less prone to confusion. A patent covers the technical details (equivalent to a US utility patent) whereas a Community design right covers the aesthetic details (equivalent to a US design patent).

The key thing about Community design rights (or design patents) is that they are extremely specific. So for example USD 633087, is (one of many) Apple design patents for laptops. Does that mean that Apple has a monopoly on clamshell style laptops and that nobody else can make a portable computer with a screen section hinged to a keyboard section. Of course not - it just means that nobody else can make one that looks too much like a MacBook.  Also check out the single patent claim - "The ornamental design for a portable computer, as shown and described." That's it. No generalisation, no fancy legal language. What you see in the pictures is all Apple get.


Likewise, the Community design right at the heart of the Apple vs Samsung dispute, only gives Apple a monopoly over rectangles with rounded corners that also happen to look like an iPad. In this case, Samsung's rounded rectangle was found to be too similar to Apple's rounded rectangle. However, as ably demonstrated by Sony, there are other ways of making a tablet computer that neatly avoid Apple's designs.

One to read

One of the great things about the internet is that you never know what gems you might find by following a link in a throwaway comment. As a case in point, I found this excellent article today on thisismynext.com. For those that don't know, thisismynext is mainly a technology news site but one of their editors, Nilay Patel, writes the odd article on patent matters as well.

I thoroughly recommend taking a look at the article and I'm definitely looking forward to Nilay's next patent piece.

- Richard

America Invents - Part 1

A number of news sites have reported that H.R. 1249, otherwise known as the Leahy-Smith America Invents Act was passed by the US Senate last Thursday (8th September) and has been sent to President Obama to be signed into law.

The America Invents Act included various provisions intended to streamline the US patent system and harmonise it with patent systems operated by most other countries around the world. Probably the most controversial part of the Act, was the move from a 'first-to-invent' to a 'first-to-file' system

Under a first-to-file system, if Alice is the first person to file a patent for an invention she will have the right to that patent. If Bob was working on the same invention before Alice but files his patent after she does, then he will be out of luck – Alice will still have the right to the patent. Under a first-to-invent system it is possible, in principle, for Bob to be awarded the right to be awarded the right to the patent, even though he filed it after Alice. Wikipedia has a helpful summary of both systems.

Moving away from first-to-invent has been widely criticised as giving too much power to large corporations to the detriment of the small inventor. Much internet commentary has focused on the supposed ease with which those corporations will be able simply steal inventions from those small inventors and then, with their greater resources, quickly file a patent, leaving the hapless small inventor with no way to challenge the theft of their inventions.

Put like that, first-to-file does seem terribly unfair. However, in my mind, this 'lone inventor vs the big bad corporation' argument does miss a couple of points.

1. First-to-invent is great in theory but not so great in practice. Any small inventor claiming the rights to a patent will need to instigate interference proceedings against the patent owner. These are basically a court case to decide the issue – and they're not cheap. I read one article (in an extremely reputable IP blog), stating that interference proceedings could cost up to $500,000. Lets assume that figure is wildly exaggerated and is out by an order of magnitude - $50,000 is still an awful lot of money for the small inventor. To me, first-to-invent seems like more of a hypothetical security blanket than a practical deterrent against that big bad corporation.

2. First to file cuts both ways. If a small inventor beats a large corporation to the patent office and files a patent for a new invention first, he/she doesn't have to worry against interference proceedings being launched against him/her, despite the fact that the large corporation would presumably be much more likely to have $50,000 lying around to do just that. And wouldn't that large corporation make a tempting target to approach for a licensing deal?

3. At the end of the day, litigation is expensive. Regardless of whether a patent is granted under first-to-file or first-to-invent, it is far easier for a large corporation to enforce their own patents against a small inventor, to challenge any patents owned by the small inventor in court or, in the worst case scenario, simply to infringe the small inventor patent, knowing that the hapless small inventor will likely not have the money to sue for infringement.

No patent system is perfect and I believe that first-to-file vs first-to-invent actually makes very little practical difference. That being the case, why not have a US patent system that is more in line with systems operated throughout the rest of the world? Having a broadly equivalent set of rules throughout the world is surely a good thing making patents easier to use for everyone, regardless of where they file.