The Program

Well, PeerNet is progressing.  I just began the programming of the Network Pool.  The network pool is a class to manage the creation of virtual sockets to connect to other peers within the network.  It also keeps track of all the peers in the network.

So, what’s been completed so far?  I’ve completed and tested the following:  Blowfish cipher, Network Socket (TCP), Encrypted Socket (bfSocket), Packet API, debugging API (including MemTest, and a logging system), and other simplistic API’s (Getting time, Packet Decoder, etc).

Here’s what I need to finish: Network Pool, the Main bootup system, and then any modules that are going to be added to the system (including a test module, chat, and statistics).

So, I hope I’ll be testing the system within the next few weeks!

The original system for a p2p filesystem was unstable, and to the point that I was unable to easily track down the bug.  Now, this didn’t call for a complete rewrite (given enough time and  effort I’m sure I could find the bug).  What made me want to do a complete rewrite was that I had some ideas on improving the system on a whole.

One of the main features of the new code is the extendability to more than just file sharing.  In example, I’d also like to add some sort of secure chat functionality to it.  This is why the name was changed, as well.  It is not longer just a file system, but a whole architecture for a LAN p2p system.  It also has a more modular design, making adding features and correcting bugs easier than the previous design.

I hope that this new rewrite is easier to find the bugs in, as well.  Though, I did just spend the last eight hours or so looking for a bug which turned out to be in a system that I thought was 100% bug free.  I guess not.  I hope to have this done by the end of the summer so I am able to test it in a real environment.

Work continues on my PeerUFS.  If you don’t know what this is, it’s a GPL’d project to take filesystems on many different computers and union them using p2p style networking code.  So far I’ve tested the project three-way locally, and it worked fine.  When I tested it two-way remotely, however, I ran into a few bugs where the program would “freeze”.  It didn’t actually go into an infinite loop or anything, it just was so slow I considered it useless (a reaction time of 120+ seconds).  I’m not entirely sure where this huge delay is coming from.  I know it has to do with data transfer timing out, but I’m not sure why.

I hope to work with some friends later to figure out the problem, but for now.. well.. ya.  I need realworld setup to help debug this program.