Thanks, DNS. It's always intriguing to see how much can be achieved with so little. Interesting article. It would have been an even better article if the author hadn't succumbed to bogus arguments, in parts. E.g.:
"Because of the staggering costs of wiring the nation's homes for high-speed networking, only 7 percent, or 7.5 million homes, now have high-speed Internet access, according to a February report from the Federal Communications Commission."
Of course, the truth is that a majority of homes now have access to such services, but only a relatively small percentage of residential users have thus far elected to subscribe to them.
====
This article and others I've read recently give off an aura of there being some form of free lunch somewhere between here and there, and I think this needs to be better understood. The neighborhood networks (nanny-nets) that I posted about in this space, several months ago or so, also fall under this umbrella, as are the aspirations of those promoting wireless meshed networking approaches, which are now receiving widespread, and imo justified, attention.
While local access to the nearest participating access point or mesh-enabled nodes can be achieved for what appears to be, or, in fact often is or will be, "free" access at the moment for some, I sense that it's inevitable that the longer term effects of displacing established forms of access (and beyond access, if the trend continues, into the next layers of service provisioning) will cause a backlash that will ultimately begin to affect how users' rate their satisfaction with network performance.
NOT because of any particular affinity that I possess for incumbent service providers (Forsooth!), but because a logical extension of this trend is to not stop with the ILECs and MSOs, but to move up the chain into the ISPs and backbone providers, as well. Whether this is a good thing or a bad thing I don't profess to know. Nor do I profess to be able to predict with certainty that it will happen. I'm just citing what appears to me to be the logical extension of a present and sustained trend that is only picking up speed, not slowing down.
Built in to the revenue streams that traditional providers receive for access services (including dialup), are subsidies that go toward their total capital- and operating-expenses. Service providers common equipment/plant assets, including buildings, vehicles, etc. - despite the hocus pocus that takes place in accounting - are amortized over the bulk of their offerings, almost universally. When you take the fastest growing parts of their networks away from them you have to expect that their remaining legacy stuff will be priced all the higher to offset those losses.
Next, if you take those legacy services away from them (think: cellular/pcs and VoIP over bootleg wireless and the resulting loss of active lines; email displacing calling volumes; work-at-homers using VPNs that support VoIP, which, at the same time supplant costly 800 number dial-in traffic to enterprise hubs that were used prior to the VPNs; the general shift in human habits in that are now more dependent on the Internet than they are on voice calling; etc.), well, then, what would appear to be the inevitable becomes clearer to see.
It's become quite fashionable to cite how the incumbents are getting their clocks cleaned by the incessant flow of innovation. But at what point in the hierarchical SP chain will the same threats that now are focused on the ILECs subside for the other SPs, of all shapes and sizes, who reside upstream toward, and into, the core?
For example [and hypothetically speaking here, only to express the point], once the ILECs have gone into liquidation, are the ISPs next, to be followed by the Internet backbone providers? If so, who will provide, never mind pay for, transit from one region to the next, and how will the flow of traffic across vaster expanses than those that are merely locally-adjacent neighborhoods be administered?
[[In fact, come to think of it, we're a lot closer to the state that I just describe, already, than my earlier 'hypothetical' disclaimer above would seem to discount the possibility of occurring. And were it not for the graces afforded by the bankruptcy courts, allowing so many restructurings in the same space to be taking place and going debt free, just to begin the cycle all over again some might argue, we'd be there in actuality, already. Quite the paradox, wouldn't you agree?]]
But getting back on track, without ISPs why would there be a need for an IETF or address and numbering admin (IANA) or an Internet Society? Certainly the ITU would be useless by that point, if the antecedents I cited above held true. And this being the case, will meshed multihop wireless networks suffice on a global scale, without domain name registries and some semblance of directory and route administration on an ongoing basis?
And ultimately, how will starter nations taking these new avenues communicate with the rest of the globe once traditional addressing and numbering plans (not to mention meet points using negotiated gateways) have been trashed due to the future non-existence of the carriers we've grown to depend on up to this point in time?
I recently responded to a similar question on another board, which, while not addressing the Etherlinx subject specifically, generally addressed the other ideas that I've cited above. Copied below, with only minor editing:
-----------
"Innovators continue to find ways round the bottleneck."
Are they really? Or, are they re-defining and adding to the number of new forms of bottlenecks that the upstream SPs couldn't give a hoot less about solving?
These make for a great level of excitement, admittedly, depending on what one is looking to achieve with their connections beyond the nearest "access point." If we examine many of these wireless workaround models we find typically that there are a number of 802.11x end users who connect to an access point that is being supported by a lone neighborhood donor participant's dsl or cable modem connection.
This donor connection is responsible for allowing each of the other participants nodes to access the Internet by way of the traditional service providers' widgetry, and this is fine for local connectivity within a geographically-local community of users. It even suffices for a limited level of casual surfing, if the number of users is not egregiously high. Think of forcefitting 16 users who are pumping 10 Mb/s each into a dsl line that is capable of running at 384 k or 768 kb/s. This might be a statistically achievable feat under casual surfing conditions, but would soon be increasingly impossible to bear with the introduction of multimedia and xbox-like apps, or even serious work at home vpn'ing.
And it may also work rather well when the access point has a robust connection to the Internet, as in a Starbucks or an Airport link to the upstream using T1s or greater speeds. The latter "legitimate" forms of hot-spotting can all be engineered to meet whatever levels of service that those entities desire, since they, unlike the bootlegger, are legitimate business entities who are presumably backed by a budget and a business case that takes these matters into account.
However, and this is more than just my humble opinion, many of our latter day "innovators" as they are now being perceived to be, will face a rude awakening when multiple users on their networks attempt to access XBoxOnline and other m-m apps as they become available, especially if more than one of them are trying to get there through a single bootlegged dsl connection being "donated to the cause" by a single end user's access line.
Consider, the person who is paying for the dsl line will find that he is being swamped, causing them to reconsider what they are doing. If the foregoing doesn't put a stop to their egalitarian ways, then the carriers and cable-ops will, through upper layer packet sniffing and the resulting blocking of multiple IP addresses and classes of service (such as VPNs and streaming content) that they will argue are not covered under residential grade tariffs. Comcast and several other MSOs have already begun doing this.
It's only a matter of time before the ILECs do the same. Thus far they've not been too aggressive in this respect, primarily because they don't face the same conflicts in the areas of content and programming that the MSOs do. Also, because the applications being accessed over the dsl lines thus far have not presented them with much of a problem. It remains to be seen what actions they will take, and if they can sustain adequate levels of service, under an onslaught by eager users of XBOX and other gaming applications, going forward.
Offsetting these last two detractors is the possibility that any number of potential innovative wireless mesh network models may emerge and begin to proliferate. These will have the effect of extending the diameter of wireless reach of end users by tying together multiple hubs and/or actual end user nodes.
But ultimately, if one wants to reach across the Internet, they must employ an ISP and backbone providers, or a number of ISPs, and those ISPs want to be compensated for "their" upstream link costs (read: T3 lines, OC3s, etc.) and other capex/opex expenditures, just like the ILECs and MSOs do.
So again, while communities might be able to sustain free access schemes for their local users, it's the business models of the ISP community and the underlying carriers who provide them with basic connectivity and transmission lines - as well as the backbone providers who span the greater distances - that must be sustained in the end. Not the innovator's local mesh, alone.
"The internet is slower.... the backbone needs are increasing just as all the component makers and carriers are going bankrupt."
Do you see the dichotomy at play here that will motivate the carriers taking the measures that I mentioned above? They will not bolster their core infrastructure to support end users who are not paying for service, imo. Instead, they will take measures to ensure that only users who are paying for service will receive service.
This is easier said than done, granted, but as someone else mentioned in another thread, the carriers will be more inclined, going forward, to begin spending more on OSS and network surveillance provisions than they would on blindly throwing more bandwidth at the problem (that would in a way continue to add to the supply of unrecoverable costs thrown at an already glutted condition), as has been suggested repeatedly by some advocates of an entirely-stupid network model. And while the SPs proceed to introduce their remedial measures, yes, agreed, throw more bandwidth at the problem. But the SPs at the same time want to make sure that the recipients of that bandwidth will be only those who are paying customers.
FAC |
