Re: The RBOCS position, Kim M. said that the RBOCS were split roughly in thirds, those strongly supporting CAP, DMT and I don't care just make up your mind. The 1/3 for CAP is what is supporting for the dual standard, which is clearly significantly short of the needed consensus for the dual standard. Kim said that the discussion at the ANSI meeting was, (I'll paraphrase) as lively.
Re: number of lines. Factors that affect lines are: distance, bridge taps, quality of twisted pair, and # of splices. All the estimates of lines that are servicable are theoretical, Kim expects that the efforts in the up coming year will give good empirical data for the numbers that we all ask for. For ADSL in general, the initial trial efforts seem to indicate that the #of lines servicable are less than originally thought and are closer to 65% than 80%. (Note: in a separate discussion that I had with a rep from Global Village (???), he said that the 65% was very doable as an economic case for the service.) Bell Atlantic's approach is oriented towards getting the % of servicable lines closer to 100%. Bell Atlantic also has the most real world experience with ADSL, i.e., they seem to understand the problems better than anyone else.
Most of the vendors I talked to were clearly favoring DMT in the long run and believed once the integrated chip sets were available CAP would go out of disfavor. There was one company who was favoring CAP who took exception to the claim that CAP was more subject to noise, but Kim M. said that was the first time anyone had ever questioned that particular characteristic. (I couldn't catch the name of the company). My sample was small and statisically subject to significant error. (let alone being heresay).
Regards,
Jim |
