dotgnu-general
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[DotGNU]Re: [Arch]Re: [Auth]a list of what we need the personal Data sys


From: John
Subject: [DotGNU]Re: [Arch]Re: [Auth]a list of what we need the personal Data system to do.
Date: Sun, 15 Jul 2001 14:31:24 -0500

> I realy don't think that we should make the decision for the people
> about who holds their information. if they want a third party to do it
> then they should have the right and the ability to do so. 

Exactly true, but by saying "a provider", you've grammatically excluded
the possibility that the customer might choose to have the ISP hold only
*certain* information, while consigning *other* information to another
provider to hold, and saving yet *other more private* information on
their own PC for self release. As described this would mean that the
customer would have have "multiple simultaneous providers" and that
DotGNU would need to support transactions using information from
"multiple, distributed providers where the customer has pre-stored
information on that provider's service".

Most providers will want to hold all the information about a customer
(ala Microsoft). Most people will even go along with this hoarding, but
we all know the risks to freedom when too much information is
concentrated in one place? Thus I say more as a matter of course: *** No
entity should hold more information about another entity in excess of
what they currently hold.***.

What they currently hold is defined by applicable laws, court cases, and
individuals decisions! The first two guarantee freedom for the third,
but if we design a system that allows for only two providers: the person
and a DataBank, we are abrogating the intent of the first two, to the
detriment of the individual!

Note: the "should" in my emphasized statement does not imply "must". The
customer is always right. If they wish to give away their freedom,
certainly we should make this option possible. However, the reverse is
also true. The customer  should be given the option to use multiple
providers to distribute their information holdings and yet percieve that
they are operating through only only one service DotGNU.

I, for one, would certainly exercise a multiple, simultaneous provider
option - my privacy is worth it. I'm certain there are other privacy
lemmings like me who will say, "I'm not stupid; I'm not expendable and
I'm not jumping off that cliff!"

> this may have
> a positive effect as it will creat compotition for  "information Banks"
> I can see subscription services that your information will not be sold,
> and I see free services that will sell your information. ISPs would also
> beable to use this as a meathod  of getting customers by saying
> Information hosting is free with a subscrition to their service and they
> will not sell your data.

True, but no customer should be locked into the single provider model.
Many may *choose* to, but let's not make that their only option. A
multiple simultaneous provider model doesn't preclude what you are
advocating. 1 is after all a multiple of 1.

Consider that having a multiple simultaneous provider model (MSPM) will
allow further competition by allowing an ISP to outsource certain
portions of their data storage, while allowing user transparency.
Further, allowing such non-centralised MSP has at least four additional
freedom yielding effects:

1) More secure even than a distributed model. If an ISP offers DotGNU
servers to its clients, and those servers contain all client
information, then that server becomes an obvious target for crackers of
that ISP. If MSP are permitted under DotGNU, then the cracker has to
guess which server at which provider contains the relevant information. 

2) MSP allow for external mirroring, at the clients option outside of
the control of the ISP, making complete DoS attacks less likely.

3) Less lock-in, a client can switch their ISP service without having to
rebuild his personal information database. He merely selects an
alternate provider, does a DotGNU transfer and he is now free to dump
his provider. Of course this imples that there must be a mechanism to
transfer data from one provider to another.

4) In MSP exclusion of mining does not depend upon a promise from the
provider. Unless the customer explicitely grants access to the miner to
all the customer's providers, no single provider can mine all the data.

Of course such distributed multiple providers will be an additional cost
to the consumer, but if we don't design in the MSP feature, we're
*inviting* ISP or other "Data Bank" to mine customer data. We're
offering the customer no *option* of protection from data mining beyond
a privacy statement from the DataBank. 

> I don't know, I Just think the option should be there for the Customer
> to make, since we are talking about freedom hear, people need to be free
> to make their own choices.

Exactly. You say "a provider" and you're taking away choice. You say
"multiple, simultaneous providers" and you're giving choice back. Now,
that is freedom.

John Le'Brecage


reply via email to

[Prev in Thread] Current Thread [Next in Thread]