Network Abuse Protections


#61

If the network has DNS, then it has a way to censor content. Whether it chooses to exercise this control is another matter.


#62

Thanks! I was rudely awakened by my 1 year old. He fell back to sleep after a few minutes of cuddles, but sadly the same could not be said for his dad! :slight_smile:


#63

Not really - an alternative DNS system could be created and could run alongside the maidsafe one. Also, the content will remain accessible via a direct address, much like IPs can be accessed directly on the clear net (although impossible to block traffic to an address with safe net).


#64

This is not the case.

DNS stands for Domain Name System, and we use the term because it describes the function, while not implying that it has the same implementation or vulnerabilities on SAFE as the equivalent system in the Internet. On the contrary.

Perhaps you are making an assumption that “SAFE DNS” is fundamentally the same as “Internet DNS” (because we use the same term in both) when they are fundamentally different, yet perform the same function? If not, please explain your reasons.


#65

Brilliant! A good read to start the day! :smile:


#66

But no one is going to want to publish their clear IP if they’re using the network for anonymity.

You can’t have alternative competing DNS’s on the same network, otherwise both server pools could resolve ‘blah.xyz’ to different addresses, and the system would fall apart.


#67

If you’ve got named addresses, they have to be translated and requests routed in order for requestors on the network to access resources on the network.

How is ‘blah.xyz’ going to serve up a website content on my storage account, to you the requestor, without some form of name resolution service on the network?

We can’t have 10 different websites thinking they’re ‘blah.xyz’ all serving up different content to different users. Otherwise, publishing URL’s as a way of providing access to resources wouldn’t work.


#68

They wouldn’t publish their clear IP. It would be meaningless if they did, as their data doesn’t live there, but in distributed form on the network.

All data has a native address which DNS maps to. This can be used directly to reach said data, if DNS is not used.


#69

Robb, I don’t have time to explain how it does this just now, and couldn’t anyway in a short reply, but I assure you it performs the functions of a centralised DNS without the characteristics of the DNS we know on the internet. See my response to your topic elsewhere on how you might go about satisfying yourself on these questions. It is well worth the effort - sorry I can’t go into detail atm.


#70

Forget IP in Safe. You must think in in XOR position. For example the hash of your top level domain is the XOR address of a list of services this name provides. And this address is unique. If somebody knows blah.xyz can find you hashing you name.


#71

blah.xyz provide abc, def, ghi

Regardless of the mechanism translating blah.xyz to a routed address (for this discussion, it’s irrelevant as to whether it’s IP, net pipes, XOR or fingers and toes), there cannot be multiple blah.xyz vanity addresses on the network, otherwise I couldn’t publish my shareable address for others to paste into the browser and access my resources - because they could be accessing someone elses blah.xyz.

Without uniqueness, the network cannot function using any type of vanity/named addresses.

The fact that there must be uniqueness means that there must be something which enforces uniqueness.

Please, when you have a moment, point me to something explaining how handles the vanity naming/claiming/resolving.


#72

The short answer is approximately “blah.xyz” is hashed (xyz is the actual name though), the hash is the XOR address of a Mutable Data with a DNS type. If you wanted to censor this name that would kinda be the same as stealing a SafeCoin, you’d have to do changes to a Mutable Data that you don’t own and don’t have any permissions modify.


#73

at the moment, claiming is realized on first arrived, first served.

My understanding is the following , please someone correct me if I’m off :

say I create coolsite.bzz

then another guy has the same idea, and tries to create it again. Unfortunately for them, the hash of “coolsite.bzz” already translates to an adress of owned data in the XOR space, and the network will refuse to create that name again.

When I point my browser to “coolsite.bzz”, the request is translated in xor space adressing, and the network hands the corresponding data to my browser.

Later, it is expected that you will have to pay some safecoin to be able to register such a name.

Regarding the unicity / non-unicity of vanity names, there have been nice discussions on the theme, including about pet naming system, which I really like :


#74

Yeah, vanity names could be a great monetisation tool, given the fact that almost every name could be re-used (unless there’s some type of intellectual property rights enforcement in the naming).

Succinct explanation - well appreciated: nice

Raises many other questions - which I’ll try to read up on before asking further, like deallocating names, how names could have their ownership transferred, etc.


#75

If you realize that the names in Safe is only a mutable data with a precise type tag, the answer to your questions is easy. Delete the MD for deallocating or MD ownership transfer to name transfer. With a little imagination, and using the ability of any MD, you can begin to understand the extreme power of this system that goes much further than the current DNS.


#76

Providing hash collisions are absolutely prevented, this seems like a sweeet approach.


#77

I feel like you do not understand the basic premise of Maidsafe.

What you’re suggesting is the exact opposite of where we are heading.
As soon as you open the network to policing you lose sight of what the network gives us: namely privacy, security and freedom.

There’s no point if my data is not my data. As soon as you open the doors to policing we end right back where we started. How could this not lead to centralisation and censorship?

Think a little bit on what you suggested.


#79

Ken when the printing press was released the media of the time thought it was demonic and that it should be destroyed. By your logic Gutenberg should have caved to the church and mainsteam and we should never have had books! Just because popular opinion is against you at the present time does not mean a technology should not be developed or released.

I honestly don’t think the mainsteam public will be our target, at least not at first. Let’s compare Facebook and Diaspora for a moment. Both run on the clearnet and both are social networks that offer much of the same features. But Facebook is vastly more popular but sells one’s data. Diaspora is more security and privacy focused however has a much smaller following. Why? Most people value convenience over privacy. It’s not that SAFE will not be valued but it’s that our adopters will be those that find the benefit of security outweighs the cost to their convenience. That’s going to be the early adopters, edge cases, outliers, outlaws, and special needs cases, not the mainsteam.