To: ftth who wrote (6026 ) 3/6/1999 1:10:00 PM From: Frank A. Coluccio Respond to of 29970
Thanks for that expansion, Dave. And thanks doubly for the p/m that I put in my weekend list of todos. I was waiting until I'd gone through it before responding to you. On that note, later. As for my reply to Greg... I had done another one of those full-blown tomes that I'm known to do from time to time, but I edited down to the main point that I was addressing. It has gotten to the point where I'm repeating myself, and decided to give myself a break. In turn, you've reintroduced some of the items I chose to delete in my unabridged version. Figures you would. <smile> But my point in my last upstream message was this: All cable modems, as well as underlying architectures, are not alike, although some of them may appear to be due to their rigid cloning from the DOCSIS standard now. There are still some renegades out there, however, who have chosen to both abide with the fringes of the new architectural rules and differentiate themselves as well. TERN, COM21 come to mind. Likewise, all DSLs and their underlying access platforms, including the means of aggregation at the central office (or pedestal), are not alike. Different DSL architectures will avail themselves to different levels of determinism in the last mile . Emphasis here meant to imply that it still makes no difference deeper into the core. Whatever else may be different between vendors products aside, the primary attributes I was addressing were the varying degrees of dedicatedness and egalitarianism that they demonstrate under duress. Your point on QoS is not lost on me, and the prospect of a new Internet is real, although the one(s) that is (are) being planned under government grants is much the same as the old in its fundamental rules set. Despite its seemingly novel appeal during the past five years, which is but a flash on the telecomm time line, and it's almost universal appeal, "The Internet" has long been a legacy platform from an architectural standpoint, however, and its primary rules set (its roots) goes back 30 years. That's longer than the concept-to-grave of ISDN and the first iterations of privatized T1. The community that the Internet serves and its servants alike find themselves going forward making maximum use (and increasing the magnification factor) of the rear view mirror. That's the way it is in this game, where Hundreds of Billions are vested in a platform that was designed to do ftp and email on a best effort basis. Only, it places a great deal of dependence now (actually it intends to leverage) the much greater amounts of fiber/lambda-enabled bandwidth that will be available on high density routes, where most Terabit routing will take place, both in the core and in the edge. This will do little (except ultimately exacerbate the situation) for the end user until similar techs are deployed in their immediate neighborhoods, and into their homes. And with few exceptions, e.g., prioritization rules such as Type of Service pointers (ToS) and class of service (CoS) weighting, the new proposed models really don't address the last mile populations all that much, something I think that is being left to the service providers to institute by way of tiered service offerings, and QoS pricing schemes, eventually. ATHM and others like UMG have already stepped into this court, with their policy edicts and bandwidth consumption-limiting rules. There will be more on this shortly, you can rest assured. ALways great to chat with you Dave, you've taught me a lot in the cable end. Later. Regards, FrankC.