




Yahoo Beats Patent Troll That Beat Google 82
jfruhlinger writes "You may recall the saga of patent troll Bedrock, which claims that it has patents over Linux and successfully sued Google over Google's Linux use. Well, the verdict from Bedrock's suit against Yahoo on similar grounds has come in — and Yahoo is victorious, not least because Yahoo went second and got to see how the arguments in the Google case went."
Re:Appeal? (Score:5, Interesting)
Just as I expected:
"First off, Bedrock had a stronger case against Google. Cawley put on evidence that Google used Bedrock's Linux code on its servers (although Google got rid of the code before trial). Yahoo, on the other hand, used a different form of Linux, and its lead trial lawyer, Yar Chaikovsky and Fay Morisseau of McDermott Will, were able to argue that Yahoo never executed the Bedrock code."
The case against Google was much stronger hence they were found to infringe. So since the cases seem to not be exactly the same, I'm guessing that Google bringing up Yahoo's case is going to mean very little to the appeal's court.
Re:Appeal? (Score:4, Interesting)
Best patent I've found? Computing the absolute value of an integer. Yes, really.
Currently owned by Oracle, previously by Sun.
Re:software patent (Score:5, Interesting)
Until (IIRC) 1986, that was the case. Software is composed of algorithms, and algorithms are by definition mathematics, and mathematics can only be discovered, not invented. So until 1986 no software patents were awarded. I think the first one was a Honeywell patent for a combined hardware/software system for an air conditioning controller.
The most egregiously bad deal about software patents is the huge list of software inventions going back to the dawn of computing, for really important stuff like virtual memory, dozens of compiler methods, the windowing GUI, many different aspects of the systems that underlie the internet, etc., that could not be patented - in retrospect those thousands of real inventors got a raw deal. I made several advances in my work in the late 1970s and early 1980s that could now be patented. All those innovations back then were either shared effectively as open source, or protected for a while as trade secrets. Now trolls can patent silly trivialities and make zillions of dollars, depending on a huge edifice of real work that they get to use for free.
And, of course, if those innovations could have been patented the entire industry would be 30 years behind where we are now.
Re:I thought Yahoo used BSD (Score:2, Interesting)
You make very good points, however I worked at Yahoo for three years and supporting three OSs was killing them. In 2003 BSD5 and Linux 2.6 were released. Assume a year for the new OS to stabilize, a year to get your internal plans together, and by 05 or so you should pick an OS and run with it. I arrived at Yahoo in '06 and found that RHEL was a second class OS to the main choice BSD4. BSD6 was just starting to show up on production machines and there was/is still tons of internally waffling about what OS was going to be the future. BSD5 never gained any traction and I think there were less a a few hundred servers with it installed.
By late 2006 Yahoo was maintaining its stack on BSD4, BSD6, and RHEL. And by stack I mean everything. Yahoo doesn't maintain an OS so much as everything else other than a base OS. However depending on the project you were on half your technology was either Linux only, BSD only, BSD4 only, or BSD6 only depending on the zealotry or skill of the team building it. Also things like yahoo-libxml moved at different paces on each OS. So you might get .17 on RHEL and .18 on BSD.
Then there were the teams that just wouldn't let go even after Yahoo had officially declared no new BSD4 servers in mid 2008. One of the largest sites at Yahoo was until last year running their Mysql databases on BSD4. And BSD4 doesn't multithread. At all. In fact Provisioning had to refuse to give them any more 8 core machines until they moved off BSD4 which still took six months of negotiating. I myself had to fight for nearly a year to move the frontend of another site to BSD6 from BSD4 in late 09. However that was Apache/PHP so we only saw a 20% increase in throughput.
In any case whether Linux or BSD, they should have picked one. If they had properly addressed this issue and picked Linux in 2005 like everyone else on the Internet they'd be in much better shape today and maintaining their stack would take half the man hours it currently does. Hell same with BSD, but frankly the mindshare of Linux and people who know how to use it still make Linux the better choice.
Lastly changing OSs is hard and it doesn't get any easier the longer your company is around. If anything Yahoo has certainly proved to me that changing even OS version is hard.