I like the idea but it seems like it works better in theory. Unless someone can really streamline the entire process (sharing, security, hardware upgrading etc) it seems ambitious that millions of people would want that kind of responsibility.
I want to work on this idea to make it into a company. I am kinda soliciting ideas from people to learn more. I think the this product will be useful precisely when all the issues you mentioned will be streamlined. A simple web based UI to the homeserver can allow for installing apps, removing them, setting up passwords, uploading data and consuming data from within the house network.
Maybe an intermediate stage, but no, not the future. Humans are lazy of nature, if they can get Facebook to host it for them (for free), why bother to buy such device and plug it in? Even if it didn’t need any setup except setting a password to manage it or so, I don’t see why anyone would do it when there are better alternatives. The internet speed vs LAN speed issue will disappear over time.
I could see IPv6 pushing this along a little more, but I agree that humans are too lazy to deal with it.
I currently have a home server, but my career is in information technology and my wife is a photographer. In my case, we have the need to have a central place to store all of our files so that they can be backed up regularly. Even being an I.T. guy, it still took me years of thinking about doing this before I ever did it. I love it now that it's in place, but when there are problems with it I'd almost rather just go back to living on the edge and storing files on our various computers than having to deal with the problems.
The homeserver can always be backed-up on an hourly basis as a fail safe. In case the home connection goes down, the backup server (maintained as a paid service by a company) can be used for a short while. I don't think traffic to the homeserver of any one family can be that huge. Torrent upload would typically have higher datarate than a regular house's upstream.
This is a horrible idea. Why would you take the job of computer security, management and purchase away from a centralized source and leave it up to the everyday laymen to handle. You say this would solve issues of privacy, but in fact would be a much less secure form of network and would likely increase the risk of security and potential data theft.
I'm a System Administrator by trade and would not ever recommend a solution like this. Centralized systems are just easier on the user and much more cost effective.
There are numerous applications that currently run on our laptops (e.g., dropbox application) that could potentially steal all your data. As we trust third party apps to behave themselves on our laptops, so can we trust the apps installed on the homeservers. With sandboxing now coming to desktop apps (e.g., thru Mac App Store), this problem will reduce even further.
This isn't about malicious applications stealing your data. It's about misconfigured or buggy applications giving people more access to your computer than you intend.
Think about how many professionally developed websites fall prey to SQL injections, CSRF vulnerabilities, server exploits and the like. Nginx recently issued a warning about a big vulnerability in all but the most recent version. Github got owned by Egor Homakov using an old Rails exploit.
And those are professionally maintained. When your average "I'm not computer-savvy, but I can work Microsoft pretty well" user is the one setting up and maintaining the server, how long do you reckon it will be before somebody takes over his computer? My guess is that you're very optimistic if your answer uses hours as a unit.
Maybe I can't tell whether the author is being ironic, or maybe I'm just old, but this is what we did in the past (10-15 years ago). Turned out to be noisy, unreliable, more expensive than you might think, and generally inconvenient, so hosting companies grew.
Computing expands, then it contracts, then it expands, then it contracts.
For every pro in any solution there's usually a similarly weighted con, both of which need to be balanced against the users technical ability, requirements, "infrastructure", etc.
There's no such thing as a silver bullet, but it's what makes the game fun.
I'd like this to happen, for all of the privacy, freedom and decentralization reasons. I think the only way it is actually going to happen is if computation speeds/prices significantly outpace network speeds.
Take for example an AI that watches your facial expressions and helps you around the house. It might be cheaper/easier to have the computation happen locally and avoid the cost of streaming all of that data to a centralized location all of the time. Then again, having all of the data in one database may provide the AI with enough benefits that it is worth sending it across the wire - with aggregate data, valuable aggregate patterns can be found. It's hard to say at this point.
I have a really hard time seeing people buying a box to support the next facebook ...
You might be interested in SourceForge's RetroShare. Direct encrypted Internet communication of all kinds between computers with no intermediate hosts.
17 comments
[ 3.3 ms ] story [ 50.4 ms ] threadhttps://wiki.debian.org/FreedomBox
I currently have a home server, but my career is in information technology and my wife is a photographer. In my case, we have the need to have a central place to store all of our files so that they can be backed up regularly. Even being an I.T. guy, it still took me years of thinking about doing this before I ever did it. I love it now that it's in place, but when there are problems with it I'd almost rather just go back to living on the edge and storing files on our various computers than having to deal with the problems.
I'm a System Administrator by trade and would not ever recommend a solution like this. Centralized systems are just easier on the user and much more cost effective.
Think about how many professionally developed websites fall prey to SQL injections, CSRF vulnerabilities, server exploits and the like. Nginx recently issued a warning about a big vulnerability in all but the most recent version. Github got owned by Egor Homakov using an old Rails exploit.
And those are professionally maintained. When your average "I'm not computer-savvy, but I can work Microsoft pretty well" user is the one setting up and maintaining the server, how long do you reckon it will be before somebody takes over his computer? My guess is that you're very optimistic if your answer uses hours as a unit.
For every pro in any solution there's usually a similarly weighted con, both of which need to be balanced against the users technical ability, requirements, "infrastructure", etc.
There's no such thing as a silver bullet, but it's what makes the game fun.
Take for example an AI that watches your facial expressions and helps you around the house. It might be cheaper/easier to have the computation happen locally and avoid the cost of streaming all of that data to a centralized location all of the time. Then again, having all of the data in one database may provide the AI with enough benefits that it is worth sending it across the wire - with aggregate data, valuable aggregate patterns can be found. It's hard to say at this point.
I have a really hard time seeing people buying a box to support the next facebook ...
what if it would be the next web?
/¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Serverless, encrypted Chat & Filetransfer
Multiple simultaneous downloads / uploads
Search Friends
Messages
Forums
UPnP / NAT-PMP port forwarding support
GnuPG Authentication
OpenSSL Encryption
Plugins support
Graphical User Interface written with Qt4 toolkit
System tray integration
\____________________
http://retroshare.sourceforge.net/