Are:
> This way I an always guaranteed that the files I share will be available
> on freenet, as long as someone can reach my node. (Using large htls).

We don't allow large HTLs because they're an easy DoS attack, and
besides, if the routing works your large-HTL request would only
encounter a relatively tiny set of nodes, and if routing doesn't
work it should be fixed not kludged around. In any case you will not
receive requests for that data except in extreme circumstances such
as tiny new networks which aren't worth considering.

Also note that monolithic files are a thing of the past. They're
going to be replaced by "splitfiles" which break a file into little
blocks and use forward erasure correction to compensate for any
missing blocks when reconstructing it. Despite the overhead of
multiple requests and redundant blocks, it's the only sane way to
deal with files of nontrivial size, and multiplexing requests is a
big throughput and reliability win. What this means is that a
"transparent store" would not only be useless, it would be
complicated.

_______________________________________________
freenet-tech mailing list
[EMAIL PROTECTED]
http://lists.freenetproject.org/mailman/listinfo/tech

Reply via email to