BitDrop (or ShadyDeliveryNetwork), a non-robotic courier system

[grue]

This system needs Faith

we can start with PGP signatures :p

[1714030702]

1714030702

[1714030702]

1714030702

[1714030702]

1714030702

[TiagoTiago]

Seems you didn't see what i did there

[Nefario]

Seems you didn't see what i did there

The opening line of the WHAM song?

I missed what you did there too.

[TiagoTiago]

http://www.youtube.com/watch?v=ocnI6hQO8SA <- watch




It was a reference to this

[Nefario]

Oh I see, the character of the game is Faith.

That is actually a cool video (seems a realistic storyline, at least from the intro), and is exactly the future we have ins store.

[eMansipater]

...is exactly the future we have in store.

I've made a career out of predicting the future--and I would never, ever use this phrase

[Nefario]

The result is each node knows only a portion of the route, it increases privacy, and the strngth of the delivery network.

Unfortunately, delivering physical objects has a very different security/threat profile from delivering data packets. All one need do to compromise the security of such a system is to send through location tracking devices. It makes no difference if delivery is done by human or non-human agents.

I was thinking about this, someone sending through a gps tracker that records the trip, but how useful would that be in reality(as long as the package is not brought to some location that is tied to the runner, i.e. their home or office or something)?

All that would say is about the pay being taken, and nothing about the person delivering.

Could lock the packages up in a faraday cage type box so they can't get a gps signal. Would wrapping it in foil work?

[Nefario]

 
Putting the object in a Faraday cage (yes, wrapping in foil) could provide a defense from live location tracking. But once the package is opened its location could be reported. So, this would provide a measure of security to the delivery network, but not to the recipient.

Interesting, I was going to say that someone would need to have the address of the recipient to send the package to them in the first place, but then I remembered the system does.

Also a package that could transmit it's location over a reasonable distance would need to be reasonably powerful and not small in size, larger than a 19inch crt monitor?

However, I do not know what you would do with a data-logging device that uses accelerometer input to reconstruct location by dead reckoning, a Faraday cage would not help in this case, it could penetrate the security of both the delivery network as well as that of the recipient. It need not even be terribly accurate, with enough samples routes can be reconstructed with a very high accuracy if they are re-used... Even completely random routes and drops would only have the effect of limiting the degree of accuracy such a attack could deliver. It could also reveal a lot of other information, including the transportation mode of the carriers, (walk/run, bike, car, etc.) as well as gait, which could be used to identify individual carriers. With enough time/space data the travel path could be also be correlated to video surveillance.

Now you're being more paranoid than I.

The best defense in this case would be to severely limit the size/weight of packages delivered, and also wrap them in foil, to prevent RFID tracking.

I agree. Any other suggestions to prevent the network being attacked?

What about law enforcement using entrapment against entry nodes? For example, they have an entry node (first person to collect package from user) collect the package which they have put illegal items such as drug into, and then proceed to arrest the entry node.

I'm thinking that you will need some reputation to be able to send packages too, this should help the network to resist.

[skittixch]

I think we're focusing too much on filling every possible security hole with measures that would stifle the proliferation and usability of such a network.  Think about the calamity that is the TSA, and the fact the people are choosing not to fly simply because the security measures have gotten so invasive. We might be venturing down a path of total anonymity, but at what cost?  I think simple, cost-effective, common sense steps can be taken (I love the tin foil packing idea) to reasonably protect all points in the chain.  If there's reason (i.e. high value, illicit, etc packages) to be paranoid about the possibility of an aggressive attack, dark runners can advertise their own methods for protecting the packages in transmission.  Instead of trying to centralize the security system, let it grow naturally out of necessity.  The runners will have reason to do this, as they'll be able to charge more for their security expertise and practices, further making a flourishing ecosystem of professional/entrepreneurial skill-sets to meet the demands of their clients.  The danger from the outset isn't outside aggression; rather, it's from setting up a system so complex that it makes people not want to try it in the first place.  Sending a package should be as easy as sending bitcoins to an acct and defining a start/end point. Any further complexity will limit the use of the system, and will run a risk of things drying up completely.  Total systemic failure is far worse than the remote possibility of an aggressive attack on a node by node basis, imo.

[FatherMcGruder]

I think that nodes should be able to specify their schedules as well as weight and size limits, but I don't think any should know the contents of the package. Otherwise, senders will simply lie about the contents. If the nodes really wanted to know, they could just peek.

If authorities compromise a node, that person can log into the site with a second password that will stop the system from routing packages through that node.

[Nefario]

skittixch you've created a monster.

I'm running with this (although I won't start until next week, kind of busy).

Also it's on Cypherpunked, I gave them an interview yesterday, should be available as a podcast in a day or so here. http://agoristradio.com/?p=285

[FatherMcGruder]

I think runners should specify whether certain routes will have privacy or not. For example, for local destinations, a runner could safely not look at packages and post his promise so as to attract more packages. However, if the runner is about to take a trip by airplane, he will have to peek at any packages that he’s carrying, and should specify that condition on his profile.

[Nefario]

I think runners should specify whether certain routes will have privacy or not. For example, for local destinations, a runner could safely not look at packages and post his promise so as to attract more packages. However, if the runner is about to take a trip by airplane, he will have to peek at any packages that he’s carrying, and should specify that condition on his profile.

That could be done.

[FatherMcGruder]

Now, there also has to be some contingency for when a runner receives a package that violates his requirements. I think the runner should just dispose of the package as he pleases.

[BitterTea]

Now, there also has to be some contingency for when a runner receives a package that violates his requirements. I think the runner should just dispose of the package as he pleases.

I was under the impression that the system would be taking care of package routing, why should the customer be penalized in that case? A runner should be able to notify the system of any exceptions (damage to package, unable to deliver, etc) and it will give them a drop off point leading the package back into the system.

[FatherMcGruder]

I was under the impression that the system would be taking care of package routing, why should the customer be penalized in that case? A runner should be able to notify the system of any exceptions (damage to package, unable to deliver, etc) and it will give them a drop off point leading the package back into the system.

I was assuming that the sender would have lied to the system for a cheaper rate. Presumably, private routes would cost more than routes without privacy.

[mp420]

I do not yet really grok the economics of this kind of P2P delivery system, but I think there might even be reasons to use it aside from anonymity concerns. There likely would be cases where rates might be cheaper than centralized mail services can offer. Centralized (public or private) mail services cross-subsidize the services to rural and other sparsely populated areas heavily. So, sending packages that are just too big to fit in an envelope might be quite inexpensive to send inside and between densely populated areas, if the couriers were people who already traveled part of the route as their daily routine (commute to work for example). Another case where marginal costs could be very low compared to conventional freight rates would of course be air travel.

This is all very fascinating, I can't wait to see how these ideas turn out to perform in real life.

[skittixch]

skittixch you've created a monster.

I'm running with this (although I won't start until next week, kind of busy).

Also it's on Cypherpunked, I gave them an interview yesterday, should be available as a podcast in a day or so here. http://agoristradio.com/?p=285

[Orbijx]

This system needs Faith

Late to the thread, but I love you so much for saying that.

Weed out the Celestes while we're at it.

[cuddlefish]

I may just integrate this into Ubitex.