On Monday 15 Sep 2003 17:39, Some Guy wrote:

> > This would not be request based. It would be a
> > polled solution. You would
> > insert a request file saying "please upload this".
> > The client application
> > polls for such requests and inserts the files. It's
> > more like automated mail
> > bots than anything else. It is at best a near-line
> > protocol.
>
> Seems like it would work, but it'd be a little slow to
> start since the data provider has to poll for your
> request first.  If his poll sits around for a while,
> and can respond later you've got about the same
> solution I was suggesting.

[...]

> Nope, I'm not try to change anybody's design or goals,
> just though I'd spit an idea out.  Once freenet's
> routing does work there may be serveral "neat" things
> that can be done with it.  The whole point is that two
> like messages can reach each other at some point in
> the network, interact and maybe return data.

[...]

> Please consider this, the freenet has an effective
> capacity that may be orders of magnitude smaller than
> the sum of all datastores.
> 1) there will be good redundancy
> 2) some data may not be found, or not by all
> At any rate C(effictive capcity) is a finite resource,
> we share.  Right now I believe the only limit on how
> much of this commons I use is my bandwidth, not the
> space I donate to the network.  Hash cache may help
> some with this.
>
> If I have 10TB of classic movies (boy I wish), I could
> never hope to insert them all into the system.  If
> they're very rare and "unpopular" they may die off
> because the one someone's looking for isn't the one I
> just uploaded.  My idea let's someone provide large
> unpopular data to the network, while only inserting
> these "advertisements" and the data as it is
> requested.  I send out say 10,000 "advertisements" and
> upload as needed.  This is much less taxing on me and
> the system in general.
>
> For the most part annoymity is there, except that I
> know somewhere out there someone is downloading.  If
> many people download the file will probably be cached
> by others and everything will be handled normally.

This all reall sounds like the job for an application level deamon. The 
node/network have the job of storing data, and IMHO, that is all they should 
do, and do it well. Everything else should live in user-land.

Gordan
_______________________________________________
Devl mailing list
[EMAIL PROTECTED]
http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl

Reply via email to