To: Shaw who wrote (1095 ) 11/22/2000 10:03:27 AM From: Frank A. Coluccio Read Replies (2) | Respond to of 2293 Hi Shaw, First off, let me just say that I'm not as up to speed as maybe I should be on AVNX's ability to deliver on some of the claims that have been made on their behalf by others. The apparent vision of the principals has been intriguing to me almost from the start, although I've sometimes had difficulty separating what has been stated by Cao from some of the interpretations of his work. Be that as it may... certainly, many of the company's visions are aligned with what many others are pursuing in optics, in principle, although many other undertakings are merely academic (the white papers are too numerous to read). And that vision centers on the creation, manipulation and management of wavelengths in a so-called all-optical manner, end to end. Simply by reading the SPIE papers that are published monthly, and copious others that have been in the mill elsewhere, I can assure you that there are a number of plays [just how many, I couldn't tell you] still to be heard from who are still doing development in stealth mode who would very much like to leapfrog this and other publicly disclosed approaches. One question that I would have is whether or not AVNX's functionality will be adopted in a broad-sweeping sense, which implies interoperability with much of the embedded base of existing plant to some degree, and if it can gain acceptance by the larger network operators, ubiquitously - versus being relegated to niche applications by autonomous network operators. To the extent that they cannot interoperate with recently-installed optical platforms, they will certainly see some formidable obstacles to their addressable market over the short term. It's the price they and others will pay for being revolutionaries. End to end means just that: end to end. It doesn't mean that you start off with a novel wavelength approach at Layer 1/2 at the originating point, and then suddenly hop onto an IP cloud and get shuffled around again under MPLS, and then return to the Layer1/2 proprietary approach at the other end of the link. I view this situation similar to the obstacles faced by RSVP historically. In order for RSVP to have worked, every router in the path had to be opted for it, with each router running at the same Rev, etc. Something, for some reason, 12,000 network operators, globally, could never come to terms with. I would welcome comments from you and others on this point. Perhaps they will be adopted first by very large enterprises and selected service providers, only. A lot will have to do with their ability make their products interoperable with others, which is no small accomplishment due to some still-very-unique aspects of their mode of delivery."To me, the virtually unlimited bandwidth capacity aspect, and speed because of no bottle necks from routers and switches, makes the system so powerfull, that demand for such a system implimented, will supply the energy to get the job done. I have confidence as a layman in Cao, based on what he has accomplished so far." The point here that gets lost sometimes is that the "virtually unlimited bandwidth capacity" quality is one that speaks to "potential", and this potential has existed since the beginning of time. It is not something that has been discovered, nor is it unique, to any one play, as some would have us believe. I like what I see here and what is being done by others, but it's still very much a learning process, as well it should be, if this space is truly going to be as disruptive as I believe it will be. FAC