---Derek Gliddenwrote:
>Most, if not all, of your ideas have been brought up
>before and
>responded to.  Sometimes politely, sometimes not.

Well, thank you for being polite.

>Essentially: freenet is not napster, gnutella or
>anything like that.

Agreed!  The problem is that *IF* you want the public
to actual use Freenet, then it needs to be easy, have
a nice GUI, and have content that people are looking
for.  Gnutella and Morpheus are big right now because
they have the content people are looking for.  

*IF* you can tap that network of data but also allow
Freenet from the same GUI, you are getting Freenet in
front of more eyes and getting another node to cache
to.

>Your best next step is to go to
>www.freenetproject.org and *carefully*
>read through the whole site and *all* the mailing
>list archives you
>can.  Most of your ideas will be debunked, refuted,
>discussed, and
>otherwise flamed upon there.

Well, they went the month of December.  :)  I have
been reading the Gnutella Developer lists, this list
and others for about a month now.  I only posted
because I am toying with the self-admittedly insane
idea of trying to tell Gnutella and Freenet together. 
I know that there are other people just insane enough
to bring this up, but when I look for them, all I find
is piles of ashes.


---(Ian Clarke) wrote:
>> After reading this list for a week, I get the
>> impression that many people see Freenet as a piece
>> of a larger application. =20
>
>This is true, one way to look at Freenet is as a
>replacement for web servers, it is an application in
>itself, but it can >be used together with a variety
>of other software (both written and unwritten)
>bringing its benefits of efficiency and anonymity.

I see and agree.  Maybe I am being too simplistic, but
it seems to me that there are four basic issues:
1) Connecting to the network.
2) Finding content you want on other nodes.
3) Download that content from those nodes.
4) Sharing my content to other nodes.

Anonymity adds another layer of complexity.
1) Connect to the network.
2) Find content without letting anyone know what I am
looking for.
3) Download content without letting anyone know what I
am downloading.
4) Share content without letting anyone know I was the
one who shared it.

Freenet can definitely play a role here.

>> I have read one post about
>> Freenet being half of a total anonymous solution.
>
>I am not sure I would agree with that, it is the
>anonymous component of a wider solution, but it comes
>with useful tools (such as fproxy) as standard, so it
>is useful stand-alone.

Yes - a wider solution!  What is that wider solution? 
Who is planning it?  Who is working on it?

>Freenet seems to search in logarithmic time in the
>average case, that is pretty much as good as you
>could hope for.

Yes, but if people starting dumping hundreds of TB of
large digital files into Freenet, how fast before
Freenet becomes bogged down?  Freenet is not Gnutella
and many of the things I hate about Gnutella are
answered by Freenet, yet *I* want anonymous Gnutella. 
I do not think I am alone.

Can Freenet play a part in this?  Is it worth
discussing using Freenet as a storage facility for the
metadata that describes what data is available for
download?

>> 2) the ability to host public files,
>
>Not sure what you mean by this.

Some nodes will feel that they have nothing to hide
and will want everything about themselves to be public
knowledge.  They will tell you anything you want to
know if it means you can download their files faster.

>> 3) some type of searchable metadata directory,
>
>As can be seen from how people currently use freenet,
>fuzzy-searching is far from a prerequisite for a
>useful system, but it would be nice.

Yes, but from one developer to another, we all know
what the end-user is and is not capable of.  Make it
too complicated and there goes your user base.

>> 4) fast downloads (caching combined with Digital
>> Fountain-type downloading?),
>
>Some clients already do multi-thread requests for
>split data, this isn't a core Freenet issue.

Agreed.  I *think* I am talking about using Freenet as
the mechanism of telling you what is out there.

>> 5) anonymous searches and downloads (with a
>> willingness to sacrifice speed and bandwidth),=20
>
>Freenet, in theory, is much more bandwidth efficient
>than Gnutella since it doesn't rely on broadcasts.

I would argue that the anecdotal evidence certain
supports your claim.

>Well, remember that our goal is not to build the next
>Napster or Gnutella, and many of the goals of such a
project would be a distraction from our core goals.

OK, I understand, but if I *WAS* trying to build the
next Gnutella, what role would *YOU* see Freenet
playing, if any?  If we had a gentlemen's agreement to
keep "large binary data" out of Freenet, what would we
be putting into Freenet?


---Sebastian wrote:
>Ian, I do think he is referring to another kind of
>search than you are. What he wants, is to search for
>e.g. "Kill AND Government" in a search engine like
>style.

That is correct.  I am looking for a way to send out
metadata in an XML format that includes a hash of the
data being described so that all XML files can be
linked back to the original data (regardless of what
you change the filename to) through the hash.

>As long as big binaries are put into *every single*
>datastore they pass on their way, I have the feeling
>that they do a lot of damage to Freenet.
>I personally see Freenet more as a document
>publishing and retrieval system (text docs, etc...)
>and not as a binary distribution means (unless the
>current behavior is changed).

I agree.  My first thoughts are to use Freenet to host
the metadata so that you never know who is hosting the
file you are looking for, but you know it is out
there.  Being able to take the hash from that metadata
and search on it in a "more anonymous" version of
Gnutella seems to have merit.

Or I could just be smoking crack.

hfw3


__________________________________________________
Do You Yahoo!?
Check out Yahoo! Shopping and Yahoo! Auctions for all of
your unique holiday gifts! Buy at http://shopping.yahoo.com
or bid at http://auctions.yahoo.com

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

Reply via email to