[mythtv-users] Zap2it Labs Shutting Down?

[R. G. Newbury]

stuart wrote:
> 
> Chris Petersen wrote:
>> Srikant Sharma wrote:
>>> Is it possible for us (mythtv-users) to convince them to share the 
>>> data with a very limited number of users?  We users can later 
>>> distribute.propagate the data using some peer-to-peer protocol. This 
>>> will alleviate the load on their servers and we can still manage to 
>>> get the listings.
>> Their main concern is that there are people making money off of services
>> that they are giving away for free.  Though I don't speak for TMS, I can
>> guarantee you that they'd never release listings data to the community
>> for distribution (it was suggested once when DD was first established).
>>  Their whole point is that they don't mind individual users getting the
>> data (hence offering it for free on zap2it.com), but if someone is
>> selling a product that uses the TMS listings, they need to pay for the
>> right to do so.
>>
>> To reiterate, bandwidth is likely not a major concern, but people
>> violating the terms of the user agreement are.  Without a way to prevent
>> it, they're shutting down the entire service.
>>
>> -Chris
> 
> What a mess.  Needs some creative thinking...  However, everything I 
> think of requires some work at the server end...  Something TMS probably 
> doesn't want to do (or should not have to do) when they are giving away 
> a service for private individual use...
> 
> Anyways...
> 
> How's about setting up a subscription service that requires each user to 
> have a unique key to decode personally encoded data.  You start out with 
> the current mythtv community.  Then, water mark the data for half the 
> uses. If the water mark shows up in commercial use, flag those users and 
> repeat the process on half the remaining users.  Repeat the process 
> until you identify the commercial user(s) and stop sending them data 
> they can decode with their key.  Eventually you will weed out initial 
> commercial users.  At such a point, start a new policy which restricts 
> creation of new accounts.  That is, something like a cooling off period 
> of say 1 week.  That way, if you get flagged you can't turn around and 
> create a new account right away.  This would also make private users 
> more protective of their keys.
> 
> Nothing is totally safe, and the above approach will take continuous 
> effort.  And, dare I say it, sounds very Big Brother'ish.  Anyone have a 
> better one?
>


Given that the computing power is available you could 'watermark' by 
seeding bad data into many more groupings than just one-half at a time. 
And the 'watermark' need not be a particularly visible piece of 
information...a 'mis-spelling' in a description for a late night movie, 
3 days out, etc. Even if a commercial user ran a diff against data from 
a different account, they would not be able to know which was correct. 
And the next download could 'correct' the error.

Geoff



              R. Geoffrey Newbury