3 ms·

Many seem to be missing the worst part of this: Services are forever tainted. And services are the most interesting part of the puzzle. Open source software

Many seem to be missing the worst part of this: Services are forever tainted. And services are the most interesting part of the puzzle. Open source software that runs on the client is nice, but frankly I'm not as concerned about the NSA having access to my desktop. Logging, parsing, and analyzing client data of this sort opportunistically probably doesn't scale. And it's easier for me to set up a system in my house where I can audit all incoming and outgoing traffic (not that I'm going to do it, but I'm confident that I could). The one exception to this is encryption software... more on this later. But it's hosted services where a lot of my interesting data lives, is structured in a way that makes opportunistic scanning much easier, and there is no way for me to audit it. There's no way for me to know that GMail's servers don't have a backdoor. Or likewise for Dropbox or Facebook or Citi or my Amazon, etc... This means I have to encrypt everything stored on any service (which is why my encryption software does need to be backdoor free). But most services store data about me where I don't control if its encrypted or not (it's not easy to encrypt voice calls from the end user perspective). The end result is I already have fatigue, and I haven't even done any work yet! I think most people are going to rely on an old story: "Two guys are walking through the woods when they run across a hungry bear. At this point one of the men quickly ties his shoes. The other man says, 'Why are you tying your shoes -- you can't outrun this bear!' To which the other man says, 'I don't need to outrun the bear, I just need to outrun you.'" Internet services are too valuable to not use. But it will be too much work for 99.9% of the population to preserve their privacy. The lack of privacy from the NSA will become the new normal. I have trouble seeing it playing out in any other way.

Amadou 13y ago
Services are forever tainted. And services are the most interesting part of the puzzle. I see it as a plus. But I've been against the very concept of such services in the first place. I think most centralization is unnecessary, that the reason it is so common today is because it is a straight-forward business plan for a company to insert itself as an intermediary between all of the users and thus extract money (and therefore efficiency) from their interactions. Facebook's the ultimate example. Their popularity is a result of people needing a place to easily host and share their photos. But the cheaper storage and bandwidth get, the less utility Facebook provides compared to a distributed system. I think we are at a point today where high-end phones could host all the content that the majority of people want to share. Combine that with a little smart caching between "friends" so that their phones on wifi can pick up the slack when the original host is cell-only and throw in a distributed hash table for finding new "friends" and the value that facebook provides users by centralization drops precipitously. If these NSA revelations help kickstart a new (but actually old) paradigm where each internet user is essentially self-hosting, then I believe that will be a great boon in the long run.
- bhauer 13y ago
  Precisely! If there is any silver-lining to the NSA revelations (and there really is not a silver-lining, I'm just using it as a figure of speech), it is that the pendulum is swinging back to self-control of data and applications. I've for several years wanted to see a post-cloud model where an application and data host I control runs singular application instances for me and all my devices are merely views on those applications [1]. I call it personal application omnipresence, and the "personal" aspect is of chief importance. I see the current configuration of Internet services--what I call the "plain cloud"--to be a frustrating diversion, and one that passively or actives suppresses the "personal" aspect of applications. As Amadou points out, the plain cloud is facile because businesses can easily be intermediaries. Oh, it's so difficult to connect an always-on, secure, and self-controlled host to the Internet to share your photos with family and friends? Don't worry! We'll handle that for you. Oh and look, we also get all of your photos and data. But don't you worry, we're good people. Trust us. For years, it has irked me that so many have voluntarily forfeited control of their data and applications out of convenience when alternative models that would provide convenience without sacrificing control are fairly easy to conceive. With equal R&D effort as the plain cloud has seen, these alternatives could be realized. I was angry when Google made their terms of service for Google Fiber disallow home servers. I imagine Amadou was as well. (I don't have Google Fiber, but I want it or something similar with reasonable TOS.) Preventing home servers is just perpetuating the current narrow vision of centralized hosting, and that notion is overdue for disruption. Obviously that's favorable to Google, but I don't really care about what's good for Google. Data centers in general should fear symmetric gigabit+ to the home. I know that without a doubt if I had fiber to my home, I would pull my servers out of my data center immediately and put a rack in my garage. I want self-hosting demystified and mainstreamed. As the pendulum swings back to self-control, anyone who is working to make "self-hosting" synonymous with today's popular verb, "sharing" is posed to earn my business. The model should be embraced by open source advocates. I even suggest Microsoft should leverage the building momentum and be the first industry titan to champion disintermediating the plain cloud [2]. [1] http://tiamat.tsotech.com/pao http://tiamat.tsotech.com/pao [2] http://tiamat.tsotech.com/microsoft http://tiamat.tsotech.com/microsoft

Amadou 13y ago

Services are forever tainted. And services are the most interesting part of the puzzle. I see it as a plus. But I've been against the very concept of such services in the first place. I think most centralization is unnecessary, that the reason it is so common today is because it is a straight-forward business plan for a company to insert itself as an intermediary between all of the users and thus extract money (and therefore efficiency) from their interactions. Facebook's the ultimate example. Their popularity is a result of people needing a place to easily host and share their photos. But the cheaper storage and bandwidth get, the less utility Facebook provides compared to a distributed system. I think we are at a point today where high-end phones could host all the content that the majority of people want to share. Combine that with a little smart caching between "friends" so that their phones on wifi can pick up the slack when the original host is cell-only and throw in a distributed hash table for finding new "friends" and the value that facebook provides users by centralization drops precipitously. If these NSA revelations help kickstart a new (but actually old) paradigm where each internet user is essentially self-hosting, then I believe that will be a great boon in the long run.

bhauer 13y ago
Precisely! If there is any silver-lining to the NSA revelations (and there really is not a silver-lining, I'm just using it as a figure of speech), it is that the pendulum is swinging back to self-control of data and applications. I've for several years wanted to see a post-cloud model where an application and data host I control runs singular application instances for me and all my devices are merely views on those applications [1]. I call it personal application omnipresence, and the "personal" aspect is of chief importance. I see the current configuration of Internet services--what I call the "plain cloud"--to be a frustrating diversion, and one that passively or actives suppresses the "personal" aspect of applications. As Amadou points out, the plain cloud is facile because businesses can easily be intermediaries. Oh, it's so difficult to connect an always-on, secure, and self-controlled host to the Internet to share your photos with family and friends? Don't worry! We'll handle that for you. Oh and look, we also get all of your photos and data. But don't you worry, we're good people. Trust us. For years, it has irked me that so many have voluntarily forfeited control of their data and applications out of convenience when alternative models that would provide convenience without sacrificing control are fairly easy to conceive. With equal R&D effort as the plain cloud has seen, these alternatives could be realized. I was angry when Google made their terms of service for Google Fiber disallow home servers. I imagine Amadou was as well. (I don't have Google Fiber, but I want it or something similar with reasonable TOS.) Preventing home servers is just perpetuating the current narrow vision of centralized hosting, and that notion is overdue for disruption. Obviously that's favorable to Google, but I don't really care about what's good for Google. Data centers in general should fear symmetric gigabit+ to the home. I know that without a doubt if I had fiber to my home, I would pull my servers out of my data center immediately and put a rack in my garage. I want self-hosting demystified and mainstreamed. As the pendulum swings back to self-control, anyone who is working to make "self-hosting" synonymous with today's popular verb, "sharing" is posed to earn my business. The model should be embraced by open source advocates. I even suggest Microsoft should leverage the building momentum and be the first industry titan to champion disintermediating the plain cloud [2]. [1] http://tiamat.tsotech.com/pao http://tiamat.tsotech.com/pao [2] http://tiamat.tsotech.com/microsoft http://tiamat.tsotech.com/microsoft