Anyone heard about this?

[Pokerface]

Is anyone heard about this?  A P2P twitter taking advantage of bitcoin network.

http://twister.net.co/

Seems quite interesting.

[Ian Clarke]

It's interesting, certainly the UI looks nice, however I have concerns about some of their design choices.

It seems they have a globally public user registration system, similar to Bitcoin's blockchain.  They're relying on a proof-of-work to prevent spamming of the user registration system, apparently it will take a few minutes of a typical user's CPU time to do this proof of work.  

If it were a requirement that we have globally unique and verifiable usernames then I think this is a reasonable way to do it, however I don't think this is a requirement.  I think Tahrir's approach, where the public key is the user, but where usernames are treated as convenient local labels for those public keys is a better approach.  

Tahrir's approach gets you away from the problem of people gathering up and squatting on usernames.  With Twister, just $40 spent on some beefy EC2 instances could quickly grab every dictionary word username.  The Twister paper claims that relying on public keys to identify users will result in user unfriendliness, however I do not believe this is necessarily the case if we are smart about our UI.  There is no reason that users must be exposed to these public keys.

They seem to have baked spam messages into the system itself, even making them an important part of the protocol.  Not sure how this benefits users.

They also seem to be outsourcing anonymity to Tor for those that want it, whereas we plan on incorporating a mixnet into Tahrir.  I think our approach is preferable here because the high-latency tolerance for microblogging is a big advantage in a mixnet, one Tor can't take advantage of.

The rest of their approach appears to be plausible, and of course they have a working prototype which is always a big advantage, hopefully one we can rectify soon.  It is good that other people are developing P2P microblogging systems, it demonstrates that there is a need for these.

Ian.

--
You received this message because you are subscribed to the Google Groups "Tahrir Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tahrir-developm...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
Ian Clarke

Blog: http://blog.locut.us/

[Kevin Wang]

You are right.  In certain aspect our approach seems works better.

Don't worry about the UI.  As long as the basic function is complete, I will work out a better-looking UI.

I am developing the Android client based on Steroids, which is a cross-platform framework working on both Android and iOS (also Windows Phone, maybe be supported in the future).  I believe I can release the first version once the Tahrir alpha is released.

--

Yours sincerely,

Kevin Wang

[Michael Grube]

I disagree that a blockchain is even the right way to go for any kind of anti-spamming measure unless you'd like the most economically powerful entity to have the loudest voice.

I really hope they are not using a line-for-line Bitcoin blockchain. I could start registering identities with one of my ASIC miners and everybody who doesn't want to or can't pay $6k to have comparable hashing power would just be SOL.

Social verification by trusted friends is probably the only way to go, but it's not even a requirement. It seems like people are just throwing blockchains into everything they can.

[Kevin Wang]

Social verification can be easily faked.  What they need is just one invitation from real world as a start point, to create an initial spam identity.  Then they can create invitation themselves, and eventually build a spam network.  But for bitcoin, at least not everyone can afford a ASIC machine.  Most spammers will think twice as they need to consider the profit.

The point is, all anti-spamming measures have their own shortcomings.  WOT could be one of the solutions. But on current stage our focus will be on the alpha release.

[Ian Clarke]

On Sun, Jan 19, 2014 at 12:31 PM, Kevin Wang <kev...@gmail.com> wrote:

Social verification can be easily faked.   What they need is just one invitation from real world as a start point, to create an initial spam identity.  Then they can create invitation themselves, and eventually build a spam network.

I think Michael is referring to a WOT model (similar to PGP - but hopefully easier to use!), and I don't think that can be easily faked in the manner you describe if it is implemented properly.  Web of trust isn't a binary "trusted" or "not-trusted".

 

 But for bitcoin, at least not everyone can afford a ASIC machine.  Most spammers will think twice as they need to consider the profit.

It is a bad thing for bitcoin that not everyone has an ASIC machine, the more people with state-of-the-art mining hardware, the better (as opposed to the current mining oligarchy).

 

The point is, all anti-spamming measures have their own shortcomings.  WOT could be one of the solutions. But on current stage our focus will be on the alpha release.

Right, although even for the first release the "rebroadcast" model used by Tahrir is a form of "web of trust", since (hopefully) low-quality messages won't get rebroadcasted.

Ian.

[Michael Grube]

If I associate a real name of somebody I know to a public RSA key and they sign off on a bad identity, I probably do not trust them as much, right? Trust doesn't need to be a binary thing. Web of trust can and does work(although then communities may censor material, but that is preferable to one entity).

My issue with using SHA-256 proof of work for spam blocking is that a wealthy entity trying to censor that network and prevent anonymity could make it impractical for ordinary  people to complete the number of hashes they need to create new identities. I haven't read any technical docs for this project and don't know that they are using SHA-256, but there is no reason to prefer that over web of trust.

[Michael Grube]

Whoops, didn't see the new messages. Apologies.