| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Geoff,
Thanks for the review.....
Geoff Thompson wrote>>
>>P1L5 The text: "Each of the four pairs supports an effective data rate of (2500) Mbps in each direction simultaneously." ...could easily be misunderstood to imply that there is a separate data path over each of the 4 pairs and that there is PHY level pair bonding as in EFM Cu. As I understand it this is not the case, rather that group coding across the pairs is being done. I would suggest just dropping the sentence unless better text can be developed. Chris DiMinico Reply>> I've utilized the language of 1000BASE-T whenever possible as the path of least resistance.
See 40.7 Link segment characteristics
"Each of the four pairs
supports an effective data rate of 250 Mbps in each direction simultaneously." Geoff Thompson wrote>>
P1L7 The sentence: "The term 'duplex channel' will be used to refer to a single channel with full duplex capability." should not be necessary as this is already a defined term in 802.3, see 1.4.106 in 2002 Further, if you look at the rest of 802.3 and ISO/IEC 11801 the use of the term "channel" refers to the full channel, not the per pair channel. Not a good idea to invent new usage on a per clause basis.
Chris DiMinico Reply>> I've utilized the language of 1000BASE-T whenever possible. I've included
the usage of "duplex channel" for consistency with 1000BASE-T.
See 40.7 Link segment characteristics
"The term “duplex channel” will be used to refer to a single
channel with full duplex capability." Geoff Thompson Wrote>>
P1L14 The text: "The cabling system used to support 10GBASE-T requires 4 pairs of ISO/IEC 11801 Class E or Class F balanced cabling with a nominal impedance of 100 ohms." (By the way, it is 802.3 convention to NOT use Greek symbols in drafts so that meaning is not lost when pasting text from drafts across to other applications in ASCII. The term "ohms" should be used. The conversion to Greek symbols is a task for the publications editor.) Should be changed to: "The cabling system used to support 10GBASE-T requires ISO/IEC 11801 4pair Class E or Class F balanced cabling with a nominal impedance of 100 ohms." Rationale: 4 pair cable is required. 2 instances of 2 pair cable (which is "4 pairs") is not assured of doing the trick. Chris DiMinico Reply>> I've utilized the language of 1000BASE-T whenever possible.
see 40.7.1 Cabling system characteristics
The cabling system used to support 1000BASE-T requires 4 pairs of Category 5 balanced cabling with a nominal impedance of 100 ohms. I'll accept your change to "4 pair" in order to eliminate the possibility of 2 instances of 2 pair....
Geoff Thompson Wrote>>
P1L19 The text: "10GBASE-T uses a star topology with Class E or Class F balanced cabling used to connect PHY entities." is not correct. It should be: "10GBASE-T uses point-to-point with Class E or Class F balanced cabling used to connect PHY entities." Since this is not a repeatered network the topology of interconnecting the bridges is way out of scope for this project. The scope of the project is a point-to-point link. Chris DiMinico reply>
I see your point but...network topologies comprised of point-to-point links
are often referred to as star topologies (distinguished from bus). Star topology is used here in
the context of deployment over Class E or Class F balanced cabling.
see 40.7.1 Cabling system characteristics
a) 1000BASE-T uses a star topology with Category 5 balanced cabling used to connect PHY entities.
Geoff Thompson wrote:>>
P1L32 The text: "segment ... and ... segment ... will provide a reliable medium." should be: "segment ... and ... segment ... will provide reliable media." -OR- "segment ... or ... segment ... will provide a reliable medium." Chris DiMinico Reply>>>
I'll change to "segment ... or ... segment ... will provide a reliable medium." Thanks.....
Chris
At 07:16 PM 5/4/2004 -0700, Booth, Bradley wrote: Dear Task Force members, |