Topic: BitMessage fork: "BitForum" Public anonymous forum (Read 3894 times)

A simplified solution would be something like this (probably not the most optimal solution)

Anyone can create a forum which will be indexed, but each forum need to attract seeders (like bit torrent), and the forum need votes in the index to rank high in searches.
Users who have access to a certain forum can vote on each level of the forum (Forum, Section, Thread, Post) and other authors.
From the protocols perspective all objects are addresses, and to vote on each object we send coded message to the address with the value -1 or +1.

Level 0 BitMessage index
addres: BF_0
An index containing BitForum addresses

Level 1 Users
value: "Gollum"
address: BF_1
User address created in the index

Level 2 Forums,
value: "BitcoinTalk"
address: BF_2
Gollum creates this forum by sending a message to the root address BF_0

Level 3 Sections,
value: "Speculation"
address: BF_3
Gollum creates this section by sending a message to the address BF_2

Level 4 Threads,
value: "Guess the price of BTC USD in end of 2014"
address: BF_4
Gollum creates this thread by sending a message to the address BF_3

Level 5 Posts,
value: "I think it will stabilize around 1000$"
address: BF_5
Gollum creates this thread by sending a message to the address BF_4
The only message type that can be sent to posts are votes, nothing will be created below this level.

User BF_12 likes this post and sends the vote +1 to address BF_5

'Thinking out loud' on this one...

In the same way Reddit allows posts to be voted UP/DOWN according to accuracy and relevancy... Is there any way that a BitForum type system can utilize the seed/peer functionality in order to prioritize and therefore speed up access to the most relevant posts only? (Obviously, access to ALL posts is necessary.)

Just wrote up a few ideas about how messaging, including forums, might be achieved over a DHT. Though it is very much a rough draft

https://bitcointalksearch.org/topic/proposing-a-new-approach-for-bitmessage-v2-298731

~the price of ice cream cones in Antarctica!

Heh... Tell me just one thing that is not national security?

reposting hacked material that involves national security might piss someone off ...no?

Just a nit-pick:  Bitmessage doesn't use a blockchain (AFAIK), which would make no sense for its purpose.  The blockchain is used to solve the problem about distributed consensus and for long-term storage of that consensus, which is not what Bitmessage needs.  Apart from the name and the use of standard cryptography algorithms, I don't think that Bitcoin and Bitmessage actually have so much in common.

Users who value anonymity don't mind slow speed when sending or reading articles.
But most users prefer speed instead of 100% anonymity
They can use 3d-party websites that synchronizes the forum-torrents to mysql-databases.

A solution might look like this:
Kademlia DHT (Distributed Hash Table) should be used for the data indexing
Bittorrent sync for the data storage
Bitmessage to post anonymous messages

Thanks for the answer.
So I understood following:
- It will be based on a blockchain like Bitmessage but the information stored could be of a higher amount and for a longer time.
Will be the speed like in Bitmessage or even slower due to the higher amount of data ?
- It will be more anonymous then Retroshare but also with less speed and lower amount of data. I guess so.
- That will be easier to use  and faster then Freenet is to be expected.  Otherwise wouldn't have much sense.
- But it will be also faster then Tor and easier to use then TBB ?

Anyway let's see it.

I would suggest that the core problems of BitMessage should be addressed before one can achieve the goals of BitForum. Namely that its ability to scale is poor and that there is a fairly large attack surface. Currently if you want to be anonymous you would be advised to use BM over Tor/I2P which defeats the purpose.

Im currently researching DHT-based overlays that are not only scalable but have been shown to offer fairly precise anonymity guarantees. This could nicely form a basis for BitForum.

Freenet and Tor is nerdy and will never be used by your mom, dad, girlfriend or grandma

Retroshare is nice and easy, but not completely anonymous and its more for a private group of friends creating a friend-network for communication and sharing of files.

The proposal of BitForum will allow anonymous authors to reach a public audience easily.
These changes will turn BitMessage into BitForum:
-Encryption of messages is optional, the message is readable by everyone but the author is still anonymous
-Index-torrent that tracks all forum-torrents (downloaded and seeded by all BitForum clients)
-Forum-torrents for each forum containing all messages (downloading and seeding is optional)

It is also of great importance that this software has an easy API so third-party developers can use the forum-torrents as a content-store. A politically incorrect news site like ZeroHedge could for example have it's articles stored in a forum-torrent, and they don't even have to know who the freelancing authors are since the authors uses the anonymous BitForum-client.

I like the idea.
But what would be the advantage over other projects with similar features (decentralized forum with hidden identity posting)
like Freenet or Retroshare ?
It would be just a different technical solution or would present some advantages also ?

The idea is really good and you should definitely get in touch with the project quixote guys!

 

What BitForum needs is two types of nodes:
Nodes keeping X days of messages for specific forums
Nodes keeping ALL messages for specific forums

Popular forums will of course be seeded, and forums with very few followers will dissapear over time when almost nobody seeds them.
Just like bittorrent.

I agree that the world definitely needs a more democratic, p2p alternative to forums. The best solutions we have at present are probably those based on Freenet or Tor, but that is a huge hassle for the end-user.

How this differs from BitMessage is that of the storage-lifetime of messages. It would be pretty annoying if our forum posts disappeared after a couple of days, a week or even a month. I like the way I can search through the bitcointalk forums for posts made in 2010 etc.

So I think this requires 'a bit more' from an overlay network than BitMessage currently does. In my opinion for it to be successful, it would be best implemented on top of some p2p-incentivised distributed data store. Nodes who support the network should be given economic incentives to do so.

Gollum pm me. I think we might be able to work together.

The project BitForum is more important than ever, innocent people are getting sued and even imprisoned just for posting links on forums!
People need anonymous public forums where they can express themselves openly without the fear of Big Brother.

It could be optional, so if you want to, you can boost the "ham score" of your message by doing the hashcash.  If you don't want to/can't, you don't.  Maybe you can even do harder hashcash to boost the score higher.  Many spammers aren't going to mind doing up to 10 seconds of CPU work for harder hashcash, so in the end maybe it doesn't help much.

Agree, hashcach may not be sufficient to prevent spammers. The solution will be the same as BitMessage for mobile users: using a centralized node. But the most paranoid forum users will of course run their own node since they cannot afford to trust any centralized entity.

Solution:
Writing posts: HashCash can be used to make it little bit harder to post spam or bullshit at the forum.
Reading posts: WoT-rating on addresses, it is up to each user to decide which posts to read or ignore.
Where should the WoT rating be saved? In a WoT-blockchain?

-Mechanism to prevent spamming (hashcash)

No. This won't work. Spammers with access to a lot of processing power, or a botnet will be able to spam. Legitimate users on a mobile phone will say "fuck it" when it takes them 5 minutes to post a message.

Hide everything by default. Use a WoT system for determining which messages to see.