broadstuff

Well that makes a change. Lodsys, instead of hiding behind lawyers, have issued a very public Q&A on their approach to iOS vendors offering in-app purchase. Even if you don't agree with software patents at all, it's refreshing to see a company like Lodsys trying to address the concerns of the tech community head on. In it they confirm the reading of their patents I gave yesterday and that it is indeed US7620565 they are seeking to charge the app vendors licensing fees to use. Indeed they have published the ballpark of their proposed license fee which they say is constrained to %0.575 of US app revenues for apps using their "invention."

While I'm impressed they have faced the tweet-blog-o-sphere head on, and applaud the fact they are seeking reasonable rather than extortionate remuneration, there are as a result of their Q&A some further points to be raised. Anyone who has read my earlier posts on software patents will know that whilst I'm highly critical of them, I don't think the concept is entirely meritless and my criticism is directed towards the current standards of examination and the fact for several years there was opportunistic in-filling of the patent database and that standards for has been allowed to pass as a novel technical effect have been nowhere near strict enough. Indeed the standards have been so derisible, the whole concept of software patents has possibly been irrevocably damaged in their eyes of the wider tech community.

So here are the comments I have against the Lodsys, after having read their Q&A (again I have to add the usual disclaimer. Though I do have some experience with Patents, I'm not a patent lawyer and you must use your own judgement rather than follow my advice. I'm not responsible for any decisions or business decisions you may make as a result of reading this):

1) The lodsys patent is a case in point. The threshold for examination for this one appears to have been inexcusably low. Looking again at the first independent claim, which practically speaking sets the footprint of this patent:

A unit, comprising: a memory; a transmitter; and a processor, coupled to the memory and to the transmitter, configured to: monitor a product for an occurrence in the product of a trigger event of a predefined plurality of trigger events, increment a counter corresponding to the trigger event upon detection of the occurrence of the trigger event, cause the display of a user interface, configured to probe for information regarding a use of the product, if the counter exceeds a threshold, cause the memory to store an input received from the user interface, and cause the transmitter to transmit the input to a server.

The problem here is that if you remove the phrase "regarding a use of the product," the claim still reads perfectly well and there is nothing technically new and that has not been seen literally thousands of times before the 2003 filing date [edit: though there are continuations which would require some study to determine the practical date which prior art needs to precede], that wasn't clearly done by a multitude of e.g. client server databases. However the phrase "regarding a use of the product" is a pure social (e.g. non technical) description of the use of information. As such adding it in makes no additional contribution to a novel technical effect. In other words, the patent is not novel and it doesn't add anything to the pre-existing state of the art. It's simply bad.

2) Given the patent is, in my judgement, bad, it is legitimate to wonder if the Lodsys strategy and apparently reasonable license terms are purely a device to avoid ever having it challenged and then, as a result, struck down.

3) I would like to understand the licensing arrangement Lodsys refer to, where Apple have licensed this patent. I wouldn't be at all surprised if it was licensed bundled with other patents in a strategic portfolio of which Lodsys is just one of the contributors. I wonder, if Lodsys had approached Apple, with this one patent in hand and asked for a license, would they have got away with anything other than being unceremoniously ejected from the building. I believe the reasonable tone they have adopted may be no more than a clever strategy based on their own understanding their patent is weak. I think the last thing they want is for it to be challenged, so they are opting to take a low sum from as many people as possible. Pointing out Apple has licensed the patent is a clever strategy for a company who may be aware of an essential weakness.

4) If they are authentically reasonable (and I don't rule out that they believe they are being entirely reasonable and don't share my view the patent is weak - everyone is entitled to an opinion - even one that corresponds with making money from others against with a demand against their preference), then perhaps they could add a clarification as to whether they believe this covers all in-app purchase. My analysis (in yesterday's post) is that it quite clearly doesn't. However they are adopting what is probably a deliberately vague line on the scope of their patent's application to existing in-app purchase apps. nowhere do they acknowledge it doesn't affect every in-app purchase design. FUD is to their advantage on this point and will result in more businesses taking licenses than need to.

As usual the tech world is frothing over software patents. In case you haven't heard, a number of sites have reported a select number of iPhone app developers have been hit by Lodsys with the claim they are infringing on one (or more of their patents). And, also as usual, the nature of the threat is overstated. This time the general consensus seems to be Lodsys have patents covering the in-app purchase mechanism and that, in troll like fashion, rather than take on Apple, they are going for the small guy, small software houses who have published apps to the AppStore.

I'm always amazed at how often tech articles about patents make generalisations, without reference to the only really significant piece of information, which is there on the web, for all to see; The patent document itself and the legally significant claims therin. Indeed out of the claims, one is most usually more important than all the rest and that is the first one. The first independent claim (independent claims are so called because they stand on their own and don't reference other claims) usually pretty much sets an umbrella over the scope of the patent. Having said that, especially in the US where the patent may contain a family of independent claims that are due to be broken out into a family of separate related patents - it's always safest to read each and every independent claim to gain a quick assessment of if the patent is troublesome. The independent claims are easy to spot, because they don't reference any other claims within the first sentence.

So I've scanned the claims of the Lodsys patents (US5999908, US7133834, US7222078, US7620565) and one of them immediately stands out as likely to be troubling to software vendors offering in-app purchase. Bear in mind, I'm not a patent lawyer, or making any claim over the quality of my reading and can take no responsibility if you take any actions based on what I am saying here (your own judgement and on your own head be it, etc, etc.) I do however have some experience with reading patent claims, so allow me to draw your focus to this one in particular, as listed on the USPTO website:

And here below, is the first independent claim. As annoying as software patents may be, put your annoyance to one side for a moment and think of it as like a linguistic puzzle. At first it will appear shockingly general, especially to technologists used to standard and well understood technology nomenclatures. But patent claims are a very carefully considered mixture of the precise with the general. When drafted properly, every particular of the phraseology makes a meaningful contribution to what is covered. Indeed read a few and most technologists soon start to appreciate the precision and care with which they are constructed.

1. A unit, comprising: a memory; a transmitter; and a processor, coupled to the memory and to the transmitter, configured to: monitor a product for an occurrence in the product of a trigger event of a predefined plurality of trigger events, increment a counter corresponding to the trigger event upon detection of the occurrence of the trigger event, cause the display of a user interface, configured to probe for information regarding a use of the product, if the counter exceeds a threshold, cause the memory to store an input received from the user interface, and cause the transmitter to transmit the input to a server.

In general the claim is the definition of a unit which displays a user interface and asks (or rather "probes" - it's always dangerous to paraphrase patent claims) the user for information regarding use of a product (and remember even a single bit response "yes" or "no" is information - remember what I said about the language, it is highly precise in it's generality) but that this occurs on detection of a trigger event and this in turn "corresponds" (deliberately vague) to a counter. So this appears to cover very generally replying to a question about the use of a product so that a server has information, presumably and most obviously to be able to take a further action such as provide an app update or additional feature (e.g, in-app upgrade).

Here the thing that stands out for me is use of the terms "increment a counter" and "trigger event". As is typical of patent claims, the phrase "increment a counter" has been carefully drafted to sound almost contingent to the detection of this "trigger event." But the very fact it is there means it is likely not to be entirely meaningless. For a start, to infringe this patent, the owner must have something that increments a counter that corresponds with a "trigger event" and if having a counter was merely an entirely contingent piece of frippery - the patent would be pretty useless. Therefore the inclusion of the phrase does effectively limit the patent. It limits by letting anything off the hook that doesn't include a counter that is in some meaningful way corresponding to the trigger event. So the use of the rather vague term "corresponds" is just that, vague; but not meaningless in terms of the legal scope of the patent. Patent lawyers always go for maximum generality whilst including only the tiniest bit of specificity required to make the "invention" stand out as distinct and novel and (importantly) as something that possesses technical effect. However it does mean that to infringe this patent, the vendor must be supplying a solution that has a trigger for the display of a user interface, with an increment counter that is in some way conducive to the triggering and indeed - as we read further - we see that the triggering is "if the counter exceeds a threshold" (though be careful, the threshold number isn't specified so may, of course, be no more than 1).

In other words, on my reading, in the context of in-app purchases, this only practically relates to solutions where the purchase is suggested as a product of the user trying to use a feature (whether entering it, exiting it, whatever) and that use being in some way counted (if only once) and then triggering a prompt (most obviously useful for prompting the user to make a purchase of the feature) when the count exceeds a threshold value.

Whilst this still appears to be highly general and very troubling, and whilst I personally suspect it may not stand-up as distinct over prior art if subject to a carefully exercise in object decomposition, it is most surely not the same thing as in-app purchase full stop. As usual the tech press reaction to this has been one of headline grabbing overreaction. Albeit a rather large indignant reaction is still in my view somewhat justified even for this more narrow case. However accuracy is important if we are to claim authority.

Practically speaking any vendors worried about Lodsys might want to consider avoiding tying the prompt for making an in-app purchase to the user's attempt to use a feature. Though I also think it is arguable that this patent doesn't cover an immediate prompt the instant a user attempts to use or exit a feature if that prompt is invoked based on a simple symbolic logic rather than through observing if a counter passes a threshold value.

I suspect - though I haven't yet investigated the apps from the vendors hit by Lodsys - they will be apps that allow the user to access a feature a number of times before prompt the user to purchase that feature if the user wishes to continue to use it. If that's the case, the easiest course of action for them to take is to stop the prompt being contingent on use of a counter and either make it a much more basic in-app purchase available somewhere in the app, or make it the product of simple logic which doesn't rely on storing or indexing a counter (though that may not for those unlucky enough to have been hit by Lodsys, prevent them suing for damages - it may help others). Note, in this latter case - even if a counter is used and the threshold set to 1 - it would be almost impossible to show clearly that the prompt is dependent on a counter without reviewing the source code (or executing costly reverse engineering), because such apparent behavior may also be achieved using a hard coded logical "if-then" condition.

Unfortunately, if my analysis is correct, it may mean Apple will decline to get involved. Though the patent is troubling, I believe it isn't one that effectively blocks the in-app purchase mechanism. I would love to see Apple hit back at Lodsys with a request for a declarative judgement. And who knows, they may take it as an opportunity to be seen to be backing the small guy and the myriad small business partners who help support their ecoysystem, as the very FUD patents like this create will surely have an impact on their business. And also they may feel they have an opportunity to make up in a small way for some of the lost goodwill over the way they have treated some businesses wishing to supply subscription content

[Edit: I've been thinking about his some more, and now I'm wondering if it is possible to entirely circumvent the Patent by implementing a logic that doesn't implement a pure counter. If the prompt to purchase a feature is triggered according to some other measurable factor, such as data volume, it might be entirely arguable, even if the data volume happens to correspond to a count - that the prompt is no longer contingent on the use of a counter. I'm fairly sure there is case law around deliberate implementation of complex features where that complexity is purely a means to avoid an obvious and more direct implementation covered by patent claims (I would certainly recommend you ask a patent lawyer if you are and app vendor thinking of taking this line), but still, it would be interesting to see if any developers can come up with useful and meaningful alternatives, especially if they are meaningful for reasons other than simply because they allow you to say "it's not a counter"]

Today The Register ran a report from Google I/O stating

Google won't open source fondleslab Android before 'year end'

I wrote a couple of comments on the article that now I wish I had published as a post in their own right. What is the real reason for the delay? We can only speculate as to the real motives of men, but in this case, I feel there is more than a little justification for adopting one position as above the others. To a large degree this is a rehash of a point I have already made (so you can save yourself some time if you have already read that article), but I think Google's latest announcement more than confirms it.

Google is making no money outside of advertising revenues and advertising "management" revenues. Sure they attribute about a billion in revenues to Android. But anyone who knows how these reports are prepared and who the audience are will know how companies take great care to present, within the bounds of legality, the picture they want to be seen and it's what's not said that is just as revealing as what is said. Google are also now coming under pressure for being, in terms of revenue, a one trick Pony. You can be damned sure they did everything they can legally do in their reports to dispel that notion. Only they can't because the figures don't lie. Here, from their Q1 2011, report is the tell-tale disclosure:

"Advertising revenues made up 96% and 97% of our revenues for the three months ended March 31, 2010 and 2011. We derive most of our additional revenues from offering display advertising management services to advertisers, ad agencies, and publishers, as well as licensing our enterprise products, search solutions, and web search technology."

I'm rephrase this to say what I think it really means:

"96-97% of our revenues are advertising with the rest coming from advertising management services only their is a negligible amount from other sources, so we will leave it vague as to how much of that 3-4% (ok, ok, how much of that 3%) is actually from non advertising related sources, such as Android preferred supplier relationship fees and licensing."

You can bet, with the heat already on in the earnings calls, if they had anything at all significant to say about Android licensing they would have said it. And even if talking about Android advertising you can be sure their accountants will have been instructed to attribute as much to Android as can conceivably, legally, be attributed to maximise as far as possible the appearance they are starting to diversify. But the analyst community isn't fooled.

Now consider Google are now number two in the massively exploding tablet market; yes that market where the PC industry is being turned on its head, and where for all Google's investment and for all the massive and growing consumer spend, they are making no appreciable money.

Meanwhile Apple are raking it in; $7.4 billion on iPad sales over the last two quarters, and that's also with the constraint of a huge order backlog and reduction in sales due to customers waiting for the iPad 2. To be clear iApple are earning more than half of Google's revenues on iPad alone! All on a product that has been out only a year. And yet there is Google, in the not to be sniffed at number 2 spot making, as near as damit, big fat zero outside of their same-old single-horse revenue model (and one that is highly susceptible to competition from the likes of Microsoft's Bing at that, at least when compared with the MS hegemony courtesy of owning a proprietary closed license OS like Windows). Now consider there is only one thing separating them from instant profit in this insanely exploding market, and that is their public backing of what, to the stock market money men, seems like the slightly hippy ideal of OSS (Open Source Software). That's not what I'm saying. But it's what the money men are thinking. And we should also be clear, it's in Google's power to go back on their pledge and to take full control and just start licensing Android as any proprietary closed source software vendor has done before them. The fall out would be interesting, and they would burn a ton of goodwill, but I say again - look at the opportunity cost of a promise made in youth that can be broken !

Now to technologists like you and me and to people who think about more than multi-million dollar bonuses, Google's stance might sound very laudable. But when you have to explain yourself and the huge $168 odd billion valuation the stock market perceives your company to have to hard thinking, shrewd calculating analysts on an earnings call, it is understandable that Larry Page bridled at being given the opportunity. And of one thing you can be sure, if Google haven't made any headway diversifying revenues, the next call will be much harder. Think of that feeling you have if there is one thing you promised your boss and you suddenly realise the next progress meeting is already upon you and it hasn't been done. That's what it's like in these earnings calls multiplied by the sums involved and the fierce torchlight of exposure in an all too public crucible.

It's becoming all too apparent to investors, the opportunity cost of keeping Android Open Source as opposed to closed and proprietary is so staggeringly huge, Google are beginning to look more than a bit, well, like a bunch of impractical idealists who don't have the balls and guts needed to create the diversified revenue streams they so badly need.

With their pledge to Open Source, they have made themselves a hostage to fortune. But confronted by a staggering revenue earning opportunity they are getting cold feet and, rather than do a full "we admit it, we're no more or less evil than the rest" about-face, they are attempting to leverage value by keeping Android closed long enough it may as well be, and are seeking to leverage money through charging for key add-ons like the Marketplace and preferred partner agreements and seem to be hoping no-one notices why they are doing it. In other words they are trying to have their cake and eat it.

So far to a degree it is working. Many of the more Fandroid of the worker bees in the development community, like brides left at the altar, are now wandering about a bit lost, and coming up with every excuse as to why their beloved Google has abandoned their ideals, except, pierced by Cupid's arrow, they refuse to see the "ugly" capitalist truth. So excuses currently include,

"The codes not good enough for public viewing." (when it's kind of the point of Open Source that community feedback patches up the difficulties)

"It's needs to be merged back into Android for phone first." (No it doesn't)

"It's to protect the open source way from predatory companies who would adopt it and fail to adopt the spirit of Open Source." (No answer required !!!)

But the simple fact is the longer Google keep it under wraps the more they can leverage money through preferred partnerships and charging for their closed source pre-integrated, advantage over the competition, key strategic technologies like the Marketplace. It's sad that devoted and energetic followers are left out in the cold by the harsh realities of unforgiving corporate finance, but that I fear, is the truth.