Ole, |-----Original Message----- |From: Ole Troan [mailto:[email protected]] |Sent: Tuesday, August 09, 2011 11:08 PM |To: DENG Xiaohong ESP/PEK |Cc: Tetsuya Murakami; Satoru Matsushima; Rémi Després; Dan Wing |Subject: Re: Port allocation considerations for 4rd | |Xiaohong, | |> Like I stated in the former mail in this thread, I'm not suggesting |> 4rd use the exact algorithm as mine, but just suggesting 4rd |algorithm |> should take 2 more points in to account, as what Remi |summarized in this thread before: |> (a) optimize chances of UPnP to ask for an acceptable port |when trying _consecutive ports_. |> (b) mitigate this requirement to have odd-even pairs in the port sets | |if I interpreted Dan Wing's comments correctly, it is not at
Yes, you made it clearly. |all clear that these should be requirements for a port mapping |algorithm. If I don't recall wrong, in the previous discussion, there are still other points, for example, legacy FTP implementations (do not extended to RFC2428) couple even-odd ports, to make this consideration open to discuss. P.S I extend this discussion to the list to lead to more comments to us, hopefully.;-) Xiaohong | |cheers, |Ole | | _______________________________________________ Softwires mailing list [email protected] https://www.ietf.org/mailman/listinfo/softwires
