[Top][All Lists]

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

Re: [Netsukuku-vala] next release of Netsukuku

From: Luca Dionisi
Subject: Re: [Netsukuku-vala] next release of Netsukuku
Date: Tue, 25 Nov 2014 11:03:51 +0100

The first documents are on the wiki and can be read and navigated
starting from here:


On Mon, Nov 24, 2014 at 12:28 PM, Luca Dionisi <address@hidden> wrote:
> The next release of Netsukuku shall have the following objectives:
> 1. make the code more modular; the modules must be able to operate
> each one independently from the others, each one just exposes a well
> described interface to the others.
> 2. document for each module:
>     * a design document (what it has to accomplish and why);
>     * the interface that the module must expose to the outside;
>     * technical implementation details.
> 3. more closely reflect the original specifications (i.e. the QSPN
> document that you can read on the official site and Aplt's master
> thesis "Scalable Mesh Networks and The Address Space Balancing
> Problem")
>     * document in detail any additional module and explain why it is needed.
> 4. implement better strategies (some new ideas) for the following
> critical issues:
>     * initial hook of a node.
>     * splitting of a g-node.
>     * migration of an entire g-node while keeping in each node its
> knowledge of routes inside the g-node.
>     * mechanism, during the migration of a g-node, to keep safe any
> open TCP connection inside the g-node itself, because the graph does
> not change inside.
>     * splitting of the network.
>     * merge two separate networks while keeping the knowledge of both
> graphs already explored.
> 5. implement new features:
>     * mechanism for any node to request to be anonymized during the
> forwarding of its packets to any destination; as a result of this, we
> get the feasibility of a peer-to-peer service that makes "anonymous
> servers" contactable.
> With reference to points 1 and 2, we wish to achieve these goals for
> the following reasons:
> A. Make possible automated and systematic testing of the functionality
> of each module of the software.
> B. Facilitate the insertion of new developers and allow the breakdown
> of tasks and doing them in parallel.
> To achieve these objectives it is necessary to rewrite almost from
> scratch most of the code. Yep, I am serious about that, but don't be
> alarmed, I swear it will be shorter and more readable than now!
> I have already started this process.
> Soon I will publish the first design documents of the single modules.
> Since those documents may require frequent changes during the
> development of the software, I will write them in Italian. They will
> be translated into English only if/when there is a
> (non-italian-speaking) developer seriously determined to contribute to
> the project.
> For the moment the documents will be published in the wiki.
> --Luca

reply via email to

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