[nspiegelberg]

Where is that open-source MapReduce stack? Facebook even gives their HPHP, Hadoop, and other code back to the community for free. The fact is, Facebook cares about Data & Google cares about technology. Google has just perfected making other companies look bad, while Facebook has the attitude of "let's keep working on our features".

[modeless]

Sorry, but if you're comparing open source contributions, Google wins by any metric. Consider Chromium, V8, Android, WebM, GWT, Protocol buffers, the Closure Javascript library+compiler+linter, and jillions of other projects (http://code.google.com/hosting/search?q=label:google), not to mention the tremendous support they've given Mozilla over the years, their work in defining and promoting open standards like HTML5, plus indirect contributions via the Google Summer of Code program.

[kingkilr]

> indirect contributions via the Google Summer of Code program.

That's indirect?!? It's basically a strait up $5 million a year investment in open source.

[kmavm]

By "any metric"? How about, "lines of open source contributed per engineer-year"? Google is 10 times larger than Facebook, and has been around for twice as long. You would expect it to have a longer list of open source contributions no matter what.

[axod]

The original article isn't about open source. It's about data portability and openness vs walled gardens.

Producing open source stuff is cool 'n all, but it doesn't really matter as much as being open vs closed in terms of what you can do with your own data.

Personally, I don't particularly care if a company decides to be totally closed source and not participate in open source contributions, as long as they don't start putting up walled gardens and trying to lock in user data.

[moultano]

Android is the largest open source release by any company ever.

[mfukar]

There's also a huge area which everyone's discounting: research. In it, Google is a beast with some major contributions and hundreds of publications, while Facebook doesn't even exist.

[swombat]

Neither company is open to releasing stuff that is, basically, its secret sauce that it makes money from.

Google makes money from their page ranking algorithm and their ad-placing algorithms, which they use to sell your attention to advertisers. You'll never see them open those up.

Facebook makes money from your personal data and connections, which they directly sell to advertisers and other partners. You'll never see them open that up.

[tobtoh]

You've missed what this issue is about and in fact you have the situation back to front. HPHP, Hadoop and other code is technology which is what Facebook is 'caring' about. Google is concerned about data - making sure that our personal data is ours to move around as we want, not when/how Facebook, Google or any other company wants.

In Google's eyes, our personal data should be portable (but under our control) and hence why they advocate openness. Google already does by allowing sites like Facebook to connect via APIs to access your contacts list and find other friends to invite to Facebook. Facebook however doesn't allow any other company to do a similar arrangement.

Google open, Facebook not. Google should be absolutely supported in their position - it will lead to be a better Net for all of us.

[rbarooah]

Except that this story is about Google becoming less open in an attempt to coerce a competitor into changing their behavior.

Google has taken away freedom from it's users to do as they want with their data (i.e. transfer it unidirectionally to Facebook). This is a fact.

So we have a clear example of Google choosing to control users data when it suits them. They claim they are doing this to try to force Facebook to become more open and that they are doing it on behalf of the users, but does anyone seriously think that's the only reason behind this decision?

Google has stated that they want access to Facebook's social graph and that they'll find a way to get it even if it's not through being granted access by Facebook. They obviously want this for competitive reasons, and to pretend that this isn't part of their motivation seems disingenuous to me.

And that's why this isn't as clear-cut as it seems. I personally think it would be good for users if Facebook's API did expose the social graph, but isn't it hypocrisy of Google to claim they are different while at the very same time holding user-data hostage?

The ends do not justify the means, particularly if the means are compromising the very principles you claim to uphold.

If open truly always wins, they wouldn't need to try to force Facebook to open up under terms dictated by Google. Google opens up data at their own pace - why shouldn't Facebook do the same?

[tobtoh]

I think it's exactly clear cut - if Google accessed some other sites data but didn't provide a reciprocal arrangement, then I would be equally harsh on Google.

'Open' doesn't win if others don't come to the party - Google is playing hardball to force Facebook to gives users control of their data. Yes they have ulterior motives but in this situation I believe the end does justify the means.

If Facebook wants to open their data at their own pace, then don't reap the benefits of 'openness', if you aren't prepared to contribute. Facebook is effectively the big selfish kid at school who only takes, but doesn't give anything back (in terms of openness).

[kmavm]

'if Google accessed some other sites data but didn't provide a reciprocal arrangement, then I would be equally harsh on Google.'

This is exactly what Google does: read the open data that all other sites provide, over http, add a lot of value by processing that data, and carefully control any 'reciprocal' access to the results of this process: namely, their search index and ad targeting systems. To be clear, I am fine with that, and you should be too. Google has added a ton of value to a lot of lives, and I am perfectly happy with them making a buck or two in exchange. That would simply not be possible if they were perfectly "open" in the way they're insisting Facebook should be.

Lots of R&D effort, some of it mine, has gone into constructing a machine-readable version of the real-world social graph. Google insisting it should have full, unencumbered, automated access to this data, because users of its GMail service have provided some inputs into it, is every bit as insane as Facebook insisting it should have full access to Google's web index, because Googlebot crawls Facebook. Perhaps Facebook should start serving empty pages to GoogleBot until Google "opens up."

Let's perform a thought experiment, stripping out the branding and associated emotional baggage. Company A and Company B both provide large scale consumer-facing services over the web, and both have large, valuable datasets that they mutually covet. There is a lot of overlap in the userbase between the two groups; users sometimes take data from Company A's service to improve their experience of Company B, and vice versa. Company A decides to shut down this access, though no users have asked them to, and though they've voted with their feet by the millions to be able to move their data from Company A's service to Company B. Is Company A striking a blow for openness against the evil, proprietary Company B? Or, is Company A just trying to negotiate a favorable market position, while manipulating your emotions the best they can?

Disclaimer: Facebook pays me to write code, not HN posts.

[mcmc]

Absolutely apples to oranges.

Google crawls Facebook's data, by request (robots.txt.) You, me, Facebook, and the world do have access to Google's complete data set... its called the world wide web.

Facebook, on the other hand, does anything it can to get a hold of private data owned by individuals, and then hoard and monetize it.

Gmail Contacts is an example of private data, as is the Facebook social graph. These are directly comparable, unlike Facebook's social graph and the greater WWW that google uses for building search indexes.

[kmavm]

In the parlance of Facebook critics, robots.txt is "opt out." But, if you don't like my web data example, consider streetview data. I don't recall asking Google to build a fleet of androids to drive around my neighborhood taking pictures of me and my neighbors' houses, but they did so anyway. They control access to this data just as closely as they do their web index. And let me re-emphasize, I am perfectly ok with that, at least until Google starts beating its chest about "reciprocity."

"Facebook ... does anything it can to get a hold of private data owned by individuals ..."

Actually, Spock, Plaxo (RIP), Rapleaf, Whitepages.com, MyLife.com, the credit agencies, governments doing background checks, etc., do anything they can to get hold of private data. They scrape online directories and listings of students at schools, digitize yearbooks and phonebooks, buy subscriber information from magazines and ISPs, buy search log data from ISPs, go to brokers who do this shady stuff and then resell the resulting information, etc. Google itself slaps a pixel on AdSense clients, and a cookie on your disk, and builds machine models that watch your web browsing history to infer your likely age, gender, education levels, interests, etc.

Facebook asks users for their personal information, and they give it to Facebook, because doing so makes Facebook's service directly, immediately better. There is a world of difference; it is the difference between surveillance and voluntary communication.

"... and then hoard and monetize it."

"Hoard"? Would you rather Facebook gave the information away to anybody who asked? Somehow that seems worse to me. Providing programmatic, API access to all of a given users' friends' personal information which is what Google is insisting on here, would be hugely irresponsible. Users have weak passwords, phishing sites exist, etc.

"Monetize": Ah, you're edging perilously close to the "Facebook sells your data!" canard. Thank you for not stating that falsehood outright. As you probably know, Facebook pairs advertisements with users that match the advertisers' criteria; the user's data is not part of the bargain. What, specifically, do you object to about this practice? Do you feel the same way about Google "monetizing" the picture it took of your home?

[tobtoh]

I don't think your Google analogy is quite the right example.

* Google indexes information that is public, that I (or others) have chosen to publish. The data on Facebook is personal data.

* Other search engines, or even yourself if you wanted to code a crawler can index that same info because there is no 'lock'. No-one besides Facebook can access my personal info (because they have put a lock on).

* This is about (one caveat - see below) being able to access my personal data and being able to port it elsewhere. It's not about ad targetting or other business aspects - ie I don't expect FB to make their ad data exportable, just my personal details.

The Caveat (which is related to your thought experiment): With regard to your thought experiment - I think the answer is both - Google's action is both a 'striking a blow for openness' and' a self-interest action to 'negotiate a favorable market position'.

[rbarooah]

You're ignoring the fact that it's the user's data - not Google's. At least according to Google's philosophy it is, and it's the users who reap the benefits.

Remember, people actually want to export their contact information from gmail rather than having to find all their friends manually, and now Google is telling them they can't because Google doesn't like Facebook's policies.

If Google believed in openness they wouldn't make that decision for their users, they would let their users decide.

Google is playing hardball by taking away data portability to their users. Hardly a good example to set.

[thwarted]

Google is effectively using the same method the GPL uses to "force" people to play nice and share. It's entirely reasonable that if someone isn't going to share with you, you don't have to share with them. Of course this takes away some freedoms, most likely the freedom of the freeloader to abuse the person sharing, but often times that's the only way to get people to understand and make things a better place for everyone.

[rbarooah]

Agreed, except that what Google isn't sharing is my contact information not theirs and so it should be my decision who I share it with. Until now Google agreed with this position and trumpeted their intention to support data portability.

Now they are only allowing me to export my data to people who play by their rules. That's not really open at all.

Sure I'd like the choice to export my contact information from Facebook, but I'd rather Google compete with Facebook on the merit of their products rather than making things harder for Facebook users.

[thwarted]

Now they are only allowing me to export my data to people who play by their rules. That's not really open at all.

False. You can go to gmail right now, click on contacts, and then choose export in the upper right. You can export your contacts all you want, then give the exported file to whoever you want. Google isn't letting Facebook access your contacts. It's not Google's fault that Facebook doesn't have a CSV importer; or maybe they do, I don't know -- either way, it's not Google's responsibility to get your contacts into anywhere else. Google isn't hoarding your information, you have full access to it.

Sure I'd like the choice to export my contact information from Facebook, but I'd rather Google compete with Facebook on the merit of their products rather than making things harder for Facebook users.

It's not Google's job to make things easier for Facebook users, that's Facebook's job.

One could say that Google is competing on the merit of their product: you can export your contacts from Google and you can't from Facebook: if access to your own data is important, then Google has the stronger product.

[arebop]

> Google has taken away freedom from it's users to do as they want with their data (i.e. transfer it unidirectionally to Facebook). This is a fact.

OK, but when you say "its users" you seem to be talking about the millions who have stored contacts data with Google, when actually you are talking about the hundreds who build apps that have other users.

Google has not taken away freedom from its users to do as they want with their data (e.g. transfer it unidirectionally to Facebook), except for those of its users that operate services or applications that limit the freedom of other users to do as they want with their data. THAT is also a fact, and one that says something about all the users who have contacts data stored with Google.

[rbarooah]

I am indeed talking about the millions of users who have stored contacts with Google.

The hundreds of application developers have absolutely no right to access that contact information.

It is the users who have the freedom to give other applications permission to access their data, and it is that freedom that Google is taking away.

To me this means that Google doesn't respect its users rights to do what they want to do with their data.

What if Microsoft prevented users from their loading word documents into Google Apps because they disliked some of Google's policies?

If Google really believes that the data belongs to their users, then they would let their users decide what to do with it, even if the users want to give it to a company that Google finds threatening.

[zaphar]

not allowing facebook to use gmail data is not the same as preventing users use their data. You can still get your gmail data and go elsewhere. All they did was restrict one company from using the API because they were using in a way that was inconsistent with Google's stated philosophy. This isn't less open it's exactly as open as it's always been for the user. The only one who has less access is Facebook.

also where do you see Google saying: "they'll find a way to get it even if it's not through being granted access by Facebook" I didn't see that anywhere.

[rbarooah]

If I want to upload my contact data to Facebook, but Google is blocking Facebook from accessing their API on my behalf, then Google certainly is preventing me from accessing my data. It is not as open as it's always been.

The point is that Google isn't adhering to their own stated philosophy in this case - they're acting as though it's their data, not their user's data.

Here's a reference to "they'll find a way to get it even if it's not through being granted access by facebook":

http://tech.fortune.cnn.com/2010/09/15/schmidt-well-pull-fac...