Re: [GNUnet-developers] Propose new feature - Search Indexer WebService for GNUnet

[SMoratinos]

Le 2012-11-07 23:21, Christian Grothoff a écrit :
> Aside from this, I should mention to SMoratinos that gnunet-java ---
> while having a pretty complete set of GNUnet APIs --- lacks support 
> for
> the file-sharing API itself AND we have no plans to fix that
Ok

Le 2012-11-07 23:21, Christian Grothoff a écrit :
> I'm just pointing
> this out as it might seriously limit your ability to realize your 
> vision.
Yes :(


Le 2012-11-07 23:10, LRN a écrit :
> "I'd suggest you to take a closer look at namespaces."
You're wright !

If I understood what you explained to me :

I send an attached file (a picture), I don't know if this is allow on
this mailling list.
The picture is to clarify the 3 entities :
 - uploader (who want publish contents)
 - downloader (who want download contents)
 - database owner (who want nothing, it's just a peer on the network,
 like all peer)

1 / The Uploader create a namespace for is content.
In the picture the namespace is "goodfiles".

2 / The Uploader choose (randomly or not) a Namespace (not own by 
himself)
in the network. The owner of the choosen namespace will be the futur
Link Index Database Owner.
The namespace choose in the picture is "a1".
The Link Index Database Owner is the Namespace choosen owner too.
This is the uploader who choose his link database owner. This point is
essential for the censorship resistant constrainst. If malicious node
attack and censor the database/namespace owner, the uploader can switch
database by choose another namespace owner, but how the downloader
discover the new namespace (it's a problem).

3 / The Uploader publish things encrypted by namespace "a1" public key,
it publish thinks to "goodfiles" namespace.

4 / The namespace owner "a1" will continuously search for publications
for namespace "a1".
This suppose to be the default functionning of gnunet, a namespace 
search
content for his own namespace.
He find the new content and index only it's link to him database.

5 / The uploader can update his content...

6 / The database owner share all link.

But how the downloader discover the "a1" namespace ? The problem is the
same as the beginning.

If there is no central server, the link index database must be in all
the network.

All peers are a Database Owner, all the peer on the network have a copy
of the Link Index Database.
When a database is updated, this update is propagated over the network.

Search became local ! This is no longer the search result which is
propagated but this is the Link Index Database.
Finally, I'm notified by all the network from updates. New contents
are indexes by all the network continuously.

The drawback is that all peers must persit this database, and have
enough disk space but few Gbits is not a problem ?
And yes Gnunet doesn't work like this now.

So what about that ?

linkindexdb.jpg
Description: Binary data