Facebook has been back in the news again (in a not-so-positive light) for recent changes to their terms of service. Specifically, the new terms no longer allow for a user to remove his or her Content and revoke Facebook’s license, prompting Consumerist headline, "Facebook: We can do anything we want with your content. Forever." The ensuing uproar led Facebook reverted to quickly revert to their previous TOS, at least temporarily.
On one hand, it’s probably not a big deal. As Charlie puts it:
As for what Facebook's going to do with all your stuff--probably nothing. I highly doubt, if I delete my Facebook account, I'm suddenly going to see all my photos being sold on some stock photo site somewhere--or some remnant version of my account that I can't get rid of. The fact of the matter is, companies don't really fare well when they do things that piss off a bunch of people--so don't expect much abhorrent behavior from Big Brother if for no other reason than it just don't make a lot of business sense."
At the same time, it makes you wonder what prompted this change. Frankly, I don’t think Facebook has ulterior motives in what they want to do with your data, but rather this is just a standard CYA legal move.
Alex recently passed along details on how Facebook is scaling photo storage on NetApp filers. They are doing an unfathomable amount of traffic here… as of last last summer, they had 6.5 billion photos and were serving almost 500,000 images per second. The I/O demands for accessing the files are so high that common file systems buckled under the load, leading them to implement "haystacks".
Each haystack is a 10GB "blob" with an index. To display any given image, a specially crafted URL is used with an index number and magic number. Because deletion is a relatively rare operation, individual photos are not actually deleted from the haystack but only from the index. Technically the photo still exists, but you just can’t get to it anymore.
Of course, this presents major problems if you removed your content and thus revoked their license to, among other things, “store, retain [or] transmit” your content. As we know, even backing up the file internally would constitute a reproduction and thus open the door for a copyright infringement against Facebook.
If you recall, they used to have a similar warning when you tried to delete your status or other item from your news feed. Though it was raised a privacy issue at the time, that too was just the limitations of an optimized architecture.
Admittedly, Facebook probably could have handled this a little better, but we all know by now how easy it for things that seem relatively straight-forward from a legal perspective to be very messy when painted in a different light.