[ppsp] 答复: next step of tracker document

[Xiajinwei <[e-mail]>]

Hi Yunfei,

Wow, your feedback is very prompt!

Yes, the Peer IP information can be used in a limited scenario, for example, all the peers are in a enterprise network and are behind the enterprise NAT, they can transmit the enterprise files among the enterprise network via their local IP addresses. But the Peer ID is mandatory and can’t be replaced by Peer IP information IMHO.

Do you mean the conclusion is removing encoding type related text from this document? if yes, will the text be moved into another document, e.g., tracker extension draft?

Thank you!


Jinwei

发件人: zhangyunfei [mailto:[e-mail]]
发送时间: 2012年9月10日 16:16
收件人: Xiajinwei; ppsp
主题: Re: [ppsp] next step of tracker document

Hi Jinwei,
    For point1, when you mention peer IP address is optional to identify the peer. Do you mean it's *feasible* or a *suitable candidate*? If it were the case, I agree with you at this point.
   For  point2, the consensus in last IETF on this draft should be "concentrating on the message* if I don't remember wrong. Regarding the encoding part, you can ask Wes and Fabio for more comments.

Thanks.
BR
Yunfei
________________________________
zhangyunfei

From: Xiajinwei<mailto:[e-mail]>
Date: 2012-09-10 15:51
To: [e-mail]<mailto:[e-mail]>
Subject: [ppsp] next step of tracker document
Hi all,

I notice there are two concerns in tracker document from IETF 84 meeting minutes, in order to accelerate the processing of this document, I’d like to show my understanding firstly. Hope we can push this work forward and get consensus as soon as possible.

1, Peer ID is global unique to identify the Peer, from this point of view, the Peer IP address is optional to identify the Peer. IMO it could be useful in a closed swarm scenario, in which the peers (both leech and seed) are behind the NAT and they can share the media content in the local domain (e.g., enterprise inside). If I am right, I suggest some text should be given to describe this scenario.

2, Encoding xml or binary experiences a long discussion, different person have different preference. One compromise is encoding and decoding XML in binary, the related context is specified in W3C Efficient XML Interchange Working Group or in ISO/IEC 23001-Part 1 ”Binary MPEG format for XML”. Therefore, encoding both XML and HTTP in binary format are implementation options. The draft can have a couple of paragraphs providing those options in terms of implementation notes.

Any comments?

Thank you!


Jinwei