[LINK] CSIRO, 'the invention of WiFi', licence fees

[Richard Chirgwin]

Roger,

The only reason I haven't written any kind of rebuttal is that to 
address each of Mullins' errors would require reproduction of pretty 
much the entire article.

I have a special and privileged position, in a way: I was aware of John 
O'Sullivan's work in radio and in chip design before this stuff ... as a 
(sorry Jan) journalist in the late 1980s or (very) early 1990s, writing 
about the chips he helped design to get the Australia Telescope at 
Narrabri to work.

(I also met David Skellern before any of this. Actually, I sat between 
him and Graeme Hellustrand on a flight to an electronics conference in 
Adelaide, listening to their mounting outrage about the State of 
Research in Australia and dodging flying coffee. But that's probably not 
germane to this discussion!)

I have a vague recollection of CSIRO announcing the patent, but I don't 
have archives of Electronics News to check...

Anyhow.

As Mark Summerfield has pointed out on Patentology, 
http://blog.patentology.com.au/2012/04/five-reasons-why-ars-technicas-savaging.html 
and 
http://blog.patentology.com.au/2012/04/story-behind-csiros-wi-fi-patent.html 
- the Mullin article ignores that the patent survived several 
re-examinations, and a judge declined to void the patent.

His demolition is better than I could have done, so I'll only throw in a 
couple of side points.

1. A lot of commenters - including Ars' editor in defence of Mullin - 
say "this is just engineering and shouldn't be patented", which is just 
silly. The basic science - the fact that radio waves behave in a certain 
way - can't be patented (or shouldn't) - patents protect 
implementations, the engineering. Not the science, which is and should 
be everyone's.

2. Mullin doesn't understand in the least degree what CSIRO is and what 
it's for. Nor does Ars understand the line between science and 
engineering. The behaviour of radio waves is a scientific fact, which 
you can't patent (note, there are exceptions to this, but let's stay out 
of rabbit holes!). When Terry Percival tells the court "CSIRO didn't 
invent" those behaviours, that doesn't affect whether an implementation 
is still available to be patented. CSIRO didn't invent the behaviour of 
radio waves; it invented a way to exploit them.

And so on, ad infinitum throughout the whole shabby piece. It's just a 
drooling, foaming-at-the-mouth, nationalistic rant.

RC

On 9/04/12 6:26 PM, Roger Clarke wrote:
> [A chap called Joe Mullin has written an article highly critical of
> CSIRO's claim to Wifi patents and its grab of nearly half a billion.
> After a large number of comments on the first article, he's published
> a second.
>
> ["Joe Mullin is a reporting fellow at the Investigative Reporting
> Program of the UC-Berkeley School of Journalism".]
>
>
> How the Aussie government "invented WiFi" and sued its way to $430 million
> By Joe Mullin | Published 4 days ago
> ArsTechnica
> 4 April 2012
> http://arstechnica.com/tech-policy/news/2012/04/how-the-aussie-government-invented-wifi-and-sued-its-way-to-430-million.ars
>
> ...
>
> CSIRO's idea of using "multicarrier modulation" to defeat the problem
> of indoor interference with radio waves was way ahead of its time.
> However, the "multicarrier" strategy was met with skepticism by a
> tech industry going in the wrong direction. They didn't even start to
> "get it" until a key, almost magical, date: January 23, 1996. That's
> when the US Patent Office granted CSIRO US Patent Number 5,487,069.
> "After that things began to change, not rapidly right away, but they
> did change," said CSIRO's attorney at the time Dan Furniss, a
> well-respected San Francisco patent lawyer (Furniss died in 2010,
> surely a blow to CSIRO's new case).
>
> The five CSIRO scientists named on the '069 patent changed the world,
> in Furniss' eyes. "CSIRO did not invent the concept of wireless LAN,
> it just invented the best way of doing it, the best way it's used now
> throughout the world," Furniss told the jury in 2009. CSIRO
> scientists had tirelessly tested hundreds of techniques until it
> found a "unique combination" that worked at high speeds.
>
> That combination involved multicarrier modulation (also called OFDM)
> as well as two techniques called "forward error-correcting" and
> "interleaving." The combination was CSIRO's intellectual property and
> deserved protection. "In some ways it [the intellectual property] may
> be the most important property we have, because it protects our
> ability to go forward in the future," said Furniss. "It protects
> innovation."
>
> When the IEEE adopted the 802.11a standard in 1999-and the more
> widely-used 802.11g standard years later-the group was choosing CSIRO
> technology. Now CSIRO had come to court to get the payments it
> deserved.
>
> ...
>
> The closest thing CSIRO ever had to a commercial product was a
> "demonstration chipset" produced by Radiata, an Australian wireless
> company formed by ex-CSIRO employees. Radiata was acquired by Cisco
> in 2000 for $295 million but turned out to be a dot-com era boondoggle
>
> ...
>
> In 2003 and 2004, the organization sent letters to 28 different
> wireless companies asking for money but was rebuffed by all of them.
> In 2005, CSIRO picked a "test case" against a small Japanese company
> called Buffalo Technology. Buffalo wasn't like US tech heavyweights,
> already used to dealing with a barrage of patent litigation. Instead,
> Buffalo stood out for its impertinence. CSIRO official Nigel Poole
> said the company had practically "picked itself" as the first victim
> by accusing CSIRO of being "swindlers." Even if other companies had
> thought it, Buffalo said it.
>
> The case was filed in the Eastern District of Texas. Advantage:
> patent-holders. Buffalo asked to change venues, but Judge Leonard
> Davis (who has overseen all the CSIRO lawsuits) wouldn't allow a
> move. The Buffalo case didn't even get to a jury. Davis ruled on
> summary judgment that the company had infringed CSIRO's patents. He
> even issued an injunction that would have banned the company from the
> US market. Buffalo appealed but ultimately settled.
>
> The Buffalo case itself probably wasn't a huge money-maker, but as a
> "test case" it worked well.
>
> ...
>
> All of the elements of the "unique combination" CSIRO proffered in
> court as a breakthrough weren't merely old by tech standards, they
> were decades old.
>
> ...
>
> Terrance Percival, spoke on the stand. He acknowledged his team
> hadn't invented any of the core elements of its wireless strategy.
> But he insisted their solution to the "multipath problem"-that is,
> interference that gets in the way of radio waves indoors-was uniquely
> successful and speedy.
>
> "We had those concepts, but we had to work out how to glue them
> together, which is a term we use," Pervical said. "There were all
> these parameters I talked about, that you had to fine tune and adjust
> to make sure you got the best possible performance out of the system."
>
> The 2009 trial ended after four days of testimony, when the parties
> settled over the weekend. The jury never reached a verdict.
>
> ...
>
>
> Responses and clarifications on the CSIRO patent lawsuits
> By Joe Mullin | Published 3 days ago
> ArsTechnica
> 5 April 2012
> http://arstechnica.com/tech-policy/news/2012/04/op-ed.ars
>
> ...
>
> The idea that any individual institution should be able to claim
> inventorship of WiFi, and make that kind of money from such a claim,
> is a big stretch. Yet, in a press release sent out this week, the
> Australian Minister for Science and Research, Chris Evans, said:
> "People all over the world are using WLAN technology, invented right
> here in Australia, to connect to the internet remotely from laptops,
> printers, game consoles and smart phones in their homes, workplaces
> and cafes."
>
> In truth, the WiFi standards were voted into existence by a world
> engineering body that had a wide array of proposals to choose from at
> each stage of the process. IEEE is an open membership organization
> and more than 300,000 engineers belong to it. Yet while dozens of
> companies made proposals to the various 802.11 committees, CSIRO
> never submitted a single proposal-not for the original 1997 standard,
> not for 802.11a, not for 802.11b, not for 802.11g, and not for
> 802.11n.
>
> ...
>
> The cost to defend a high-stakes patent claim through trial now
> averages $4-5 million [only?] , and that doesn't include the risk of
> massive damages for a losing defendant. It's just too expensive and
> risky to defend all the cases. It's become part of the cost of doing
> business in the US tech sector. Nearly all settlements just represent
> expedient decision-making-not a comment on truth claims.
>
> ...
>
> [Mullin does overlook, or perhaps worse misrepresent, an important
> aspect of patent law.
>
> [It's completely irrelevant whether the patent-owner ever offered to
> assist a standards-setting organisations when it was developing a
> standard.  And it's completely irrelevant at what stage the
> patent-owner approached the organisations that implemented the
> standard in products, demanding a licence fee.
>
> [That may be a dreadful feature of patent law and practice;  but
> that's how it is.]
>
>