As with other commenters here, I completely agree.
The NSA/Verizon issue has been conflated with the PRISM issue, because they were both published very close together and involve the NSA and broad scale data snooping.
I am horrified about the NSA/Verizon thing - because there is a court order that explains exactly what is going on, and it doesn't require too much technical knowledge to understand how bad it actually is.
The PRISM thing will scare me if it turns out to have "direct" and "unilateral" access to the servers of these companies. However, I doubted this from the beginning. Although I agree very strongly with this article, even they don't go far enough. The term "the companies central servers" keeps coming up, but even that is not a good representation of how the companies would store data of interest to the NSA.
Do the Washington Post journalists think that you just turn on the server like you do your desktop, open a folder called "John Smith", and then see a text document with all of their chat messages? Most software developers would have enough trouble trying to figure out exactly how the DB and Application layer of a piece of software works together in the first place - some decisions that DB designers or application developers (including myself) are horrifyingly baffling. And each company would have had engineers who made equally baffling, but extremely different design decisions with regards to data storage.
Then there is the number of servers involved. Would somebody at the NSA say "Hmm, I want to get details on John Smith, so I'll just log into Facebook's central servers and pull the details up." - there would be thousands of servers involved, and plenty of duplication and de-normalization. But still, what servers have details on John Smith? There probably isn't even somebody at Facebook who could tell you, let alone somebody accessing a supposed system like PRISM.
As the original article states, I think that it is more likely a system to deal with warrants and getting info from these companies through the typical channels that most of us are probably familiar with.
Now, if there is another AT&T-esque issue where the government is tapping into fibre to store data in communication, possible for future decryption, then that is a whole nother issue. As a previous HNer mentioned (can't find the reference now) we all "know" that the government is probably doing that, but none of us KNOW that they are, until we see a court order like that from the Snowden leak.
The NSA/Verizon issue has been conflated with the PRISM issue, because they were both published very close together and involve the NSA and broad scale data snooping.
I am horrified about the NSA/Verizon thing - because there is a court order that explains exactly what is going on, and it doesn't require too much technical knowledge to understand how bad it actually is.
The PRISM thing will scare me if it turns out to have "direct" and "unilateral" access to the servers of these companies. However, I doubted this from the beginning. Although I agree very strongly with this article, even they don't go far enough. The term "the companies central servers" keeps coming up, but even that is not a good representation of how the companies would store data of interest to the NSA.
Do the Washington Post journalists think that you just turn on the server like you do your desktop, open a folder called "John Smith", and then see a text document with all of their chat messages? Most software developers would have enough trouble trying to figure out exactly how the DB and Application layer of a piece of software works together in the first place - some decisions that DB designers or application developers (including myself) are horrifyingly baffling. And each company would have had engineers who made equally baffling, but extremely different design decisions with regards to data storage.
Then there is the number of servers involved. Would somebody at the NSA say "Hmm, I want to get details on John Smith, so I'll just log into Facebook's central servers and pull the details up." - there would be thousands of servers involved, and plenty of duplication and de-normalization. But still, what servers have details on John Smith? There probably isn't even somebody at Facebook who could tell you, let alone somebody accessing a supposed system like PRISM.
As the original article states, I think that it is more likely a system to deal with warrants and getting info from these companies through the typical channels that most of us are probably familiar with.
Now, if there is another AT&T-esque issue where the government is tapping into fibre to store data in communication, possible for future decryption, then that is a whole nother issue. As a previous HNer mentioned (can't find the reference now) we all "know" that the government is probably doing that, but none of us KNOW that they are, until we see a court order like that from the Snowden leak.