[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [RAI] Draft on P2P architectures

Hi Enrico,
thanks for your comments. Yes, you are right; that sentence could be misinterpreted. What about changing it to this one instead? 

"Note, however, that a particular swarm where most endpoints were
infrastructure nodes that had the complete file from the
beginning and, thus, acted all the time as seeders could not be
strictly considered a P2P system because most endpoints would
only be providing services, not requesting them."

Cheers,

Gonzalo

Enrico Marocco wrote:

Hi Gonzalo,

the draft seems in pretty good shape and I think it should be published
as soon as possible to the benefit of the whole community. However, I'm
not sure I totally agree with the last sentence of the newly added
section 2.4, "Applying the P2P Definition to BitTorrent":

   Note, however, that a particular swarm where most endpoints were
   seeders could not be strictly considered a P2P system because most
   endpoints would only be providing services, not requesting them.

In the common case a seeder is a peer that has completed the download,
but still has not uploaded enough to reach a decent share ratio. So, the
fact that a swarm has a majority of seeders is more likely to indicate
that consumption of resources is significantly faster than provision
(either because of asymmetry in uplink and downlink bandwidth, or simply
because offer exceeds demand) and not necessarily that the peers are
providing the service without getting anything back -- they have already
got it.

Enrico

Gonzalo Camarillo wrote:

Hi Dan,

thanks for your comments. Answers inline:

Dan Wing wrote:

the IAB has just submitted the following draft:

http://www.ietf.org/internet-drafts/draft-iab-p2p-archs-00.txt

Comments are welcome.

Please add Gnutella, BitTorrent, and Skype to section 2.  I
noticed BitTorrent isn't mentioned in the document at all.
I have added BitTorrent to Section 2. However, I would not like to have too many examples in that section. Since Skype and Gnutella were already mentioned somewhere else in the document, I have not added new subsections with them in Section 2. 


You might also mention Octoshape's don't-look-under-the-covers P2P video
streaming application,
http://arstechnica.com/web/news/2009/02/cnn-p2p-video-streaming-tech-raises-qu
estions.ars


I have added references to Octoshape.


end of Section 5.3,
     traditional SIP, which relays on a rendezvous server infrastructure.
                            ^^^^^^
                            relies

Fixed.


Section 6,
   we have discussed a number of
   perfectly legitimate applications that have been implemented using
   P2P.

I would drop "perfectly" from that sentence.  (Same phrase is in Section 1).
(Are there imperfectly legitimate applications?)
I would like to stress the fact that there are *fully* legitimate. If you have a suggestion to give more emphasis to the sentence without using "perfectly", let me know. 


To your list of legitimate uses of p2p technology in Section 6, you might also
consider adding the "unofficial release" by Nine Inch Nails of 400Gb of
time-synchronized multiple-camera HD video footage from three concerts using
BitTorrent, http://forum.nin.com/bb/read.php?52,378166.  The trackers are on
nin.com's own website.
Since this is yet another example of content distribution using bittorrent, I do not think we need to add it to the document... but thanks for the reference anyway. It is interesting! 

Thanks for your comments,

Gonzalo
_______________________________________________
RAI mailing list
RAI at ietf.org
https://www.ietf.org/mailman/listinfo/rai