Google should open source what actually matters: their search ranking algorithm

It's perfectly understandable. Lots of money is made in the existing paradigm and so many of you have incentive to believe it's the most rational system. Besides, hard to imagine Google's doing something wrong, right?

But the argument that "the only people this will help is spammers"? Really? Are the only people helped by open cryptographic protocols those who look to decipher traffic?

"They have every right to keep it secret and profit off of it." Let's allow this for a moment. We've still got a false dichotomy. Making it public will not deprive them of their profit. First, incredible marketing coup. Does anyone else remember how much of Google's initial strength came from its rabid fans? This would do a lot of re-energize that base.

Second, are there going to be new startups which can effectively compete with Google, even given its algorithm? How are they going to provide a better experience, or serve the volume of traffic, or index a larger amount of pages with the same algorithm with fewer resources?

Third, who says they have to authorize competitors to use it? This is I think one of the more interesting points. They could release the algorithm but license it only for non-commercial use for anyone not themselves. Ridiculous? Because Microsoft would violate such a deal anyhow? Because the algorithm could simply be reimplemented? Perhaps. But it would be another roadblock.

Fourth, why do we continue to believe that Google's strength comes from proprietary code? Why do we not recognize that its strength comes from mindshare, user experience, and quality of execution? None of which would be negatively affected by releasing the source.

Perhaps Google would face heightened competition as a result of releasing its algorithm. Perhaps this would outweigh the benefits of doing so. But there would be significant good which would result, including public feedback which would make it significantly harder for spammers to be successful.

But I'm just a foaming-at-the-mouth open source radical. And I'm sure history is on your collective side. If there's anything we've learnt from a hundred years or so of Computer Science (overestimate if you only count actual programming, underestimate if you allow Babbage as you should), it's surely that secrecy is what drives innovation, right?

I don't know what could be done as verification for the advertiser, but the publisher should be able to do something, right?

Increasing birth rates enables economies to grow without becoming stagnant and old, placing undue demand due to pensions. An exceptionally high birth rate creates food production problems, housing problems, and generally the 'sardine can' effect. Low birthrates create a stagnant civilization that will eventually reach a tipping point when taxes get so high to pay for pensions that the younger generations will protest.

Market freedom to the extreme leads to monopolies taking advantage, 'gentlemen's agreements' between companies to price fix and many other problems that rules and regulations try to control. Equally a controlled market develops problems in and of itself as it's unable to respond naturally to market fluctuations, and quickly becomes unruly and unstable, and as many of the countries who have held long-term controlled markets are testaments to, they develop massive black-market economies to deal with these.

The zealous burning of fossil fuels, aside from global warming concerns, can expel vast quantities of radioactive materials, sulfurous emissions and a long list of things we've accepted in western society as being more harmful than smoking to the wide populations health, but rules and regulations on air quality increasingly get more lax (at least they have been where I live). With zero fossil fuel production, however, our civilization would collapse (at least with an immediate stop to production), and I'm not just talking the halting of transportation vehicles, I'm talking the stop in production of the advanced polymers used in virtually all aspects of our lives. The majority of people wouldn't have clothes on their back . . . my wardrobe would certainly be a lot cheaper if my clothes were all genuine cotton and wool. Oh and farewell cheap elasticated sweatpants.

Every good idea taken to its extreme is a bad idea and will create monumental problems in its wake. Electric vehicles are amazing, however a pure-electric is a ridiculous prospect in winter when the capacity hits below 60% of room-temperature, suddenly your vehicle can't get you to work and back on a single charge. Zero integration of hybrids is, as everyone knows, damaging to the environment but also an extreme waste of entropy. A 1 tonne vehicle at 100km/h is a massive store of energy, and allowing that to be wasted is criminal, the amount of energy humanity has wasted on braking at highway speeds would likely run the entire human civilization for over a year.

Without proof that is nothing but handwaving, if they are as impartial as you suggest they are they've got nothing to lose by letting you know what their take is and how big a portion of their clicks are confirmed fraud.

As some have incorrectly asserted, Google's latest post is not about "everything open", it is an internal memo to product managers that it occurred to them would be fit for public consumption.

The central thesis of the letter is that "open" is hard to define (and defined differently by different people) and yet crucially important. Because it is hard to define, many people misuse it (OpenDNS comes to mind), and no one can call them on it because it is a vague term.

Internally, Google employees argue about what open means, and the purpose of this letter was to clarify what open means to Google. They lay it out in extremely clear terms, and it is well structured.

This response essay by Dixon misses all that completely - he unilaterally defines "open" to mean "open source", and then asks for Google to give away their software, claiming that if they don't, they're hypocrites.

Going back to Google's definition of "open" (which is complex), one significant aspect they emphasize is the notion of "lock-in", i.e. locking in users, and locking out competitors. This is one area where Google puts their money where their mouth is; they let users leave. You can get at your email, export your Google Reader content, download all your Google Docs, archive your Picasa pictures, and export all your contacts. That means if a competitor comes along with something better, you can start using it. Try that with Facebook.

Dixon acting like that is worthless because it isn't the "open-source" form of "open" is patently absurd. If this is confusing, I highly recommend reading Simon Phipps essays on open source to understand what kind of software makes a good candidate for a business to open source (operating systems, browsers, compilers...the basics that we all use to compete) and what does not (grid/cloud computing algorithms, specialized software, and other types of software that are differentiators). Essentially, the business model is to pool our collective resources in areas in which we don't want to compete (web server software, operating systems) and to spend time innovating in areas we do wish to compete (search algorithms, voice recognition, image analysis, etc.) Over time, as those specialized areas become more mainstream, they become better candidates for open-sourcing.