[Blocks-development] Future Direction For Blocks

[Blat F. <pet...@ho...>]

Its beena bit quiet on the list recently, and that cant be
such a bad thing given the number of changes which were
released recently. V0.15.1 seems to be holding up remarkably
well, and the new uploading and readvertisement logic doesnt
seem to have any serious problems.

Its probably worth discontinuing V0.14 support, so I would
suggest Luke upgrades his server to V0.15.1 (and changes the
port number) in preperation for V0.16

V0.16 will still be released in a week or so, but the only
significant changes will really be the addition of 'sticky'
files, and upgraded Makefiles so that Linux builds cleanly
without alteration of any files.

In the meantime I was thinking about the future direction
of Blocks. There has been a lot of discussion around Michaels
Web-Of-Trust idea, but so far there is no real substance
availale for general discussion. The simulator is a good
idea for proving that Blocknets are scalable, but is a
(important) side issue to where we are actually going with
all this.

So, what do we want to do with Blocks?

Heres a few ideas off the top of my head...

1) Allow (optional) HTTP access to servers, with search and
   download ability. Perhaps upload as well. This would allow
   modemers and people without always on inet connections
   access to content on the Blocknet. If this were commonplace
   perhaps the Blocknet would simply become a fancy backbone
   to this larger systems supporting thousands of low bandwidth
   users?

2) Aim at giving good support for *huge* files. Currently
   Blocks is inefficient for files < ~150Kb anyway, but there
   seems to be a lack of file sharing utilities that are
   particularly suitable for large files (video, CD images, etc).
   This would require Blocks servers to have even larger chaches
   (at least 4+ Gb), and probably mean that files may have to
   be split across more than one server... creating an informal
   mixnet. It might be possible to simply split the large
   files into smaller ones and make Blocks simply join them
   back up on download.

3) Add a rating/voting system. It could be possibly to broadcast
   'issues' to vote on, and also allow people to broadcast 'votes'
   on these issues and also on files as well. Now, because of
   the dynamic nature of the Blocknet all the servers would have
   slightly different results, but perhaps this woudlnt really
   matter. Flooding/Spamming would seem a real issue here, but
   is it really that much worse than the general case?

4) Add TCP/IP socket routing via the Blocknet. This would allow
   IP anonymous surfing, irc, telnet, etc. This would require
   Blocks to use a smaller data block size and therefore
   multiple header blocks etc, so that delays were minimised
   for near realtime applications. This would be difficult, but
   not impossible.

Just a few ideas to get us all thinking. Please feel free to
comment or suggest new ideas. Its brainstorming time, so even daft
ideas are fine :-)

ttfn

PG.

________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com