There’s a big difference between forwarding Skype at best-effort priority and blocking it. In the US, we have yet to see an ISP block Skype, although it would be perfectly legal to do so and has been forever.

It’s also facile and misleading to suggest that ISPs inherently have the resources to give higher than best-effort service away for free, or to suggest they have some obligation to do so. If higher-priority services are only available at a higher fee, that’s a perfectly reasonable way to run a business and nobody needs to be chastised over it. And it doesn’t mean “some web sites load faster than others”.

The neutrinos seem to miss the fact that high priority only works as long as it’s not overloaded, and opening the quick-check lane to everybody defeats its purpose.

Ironically, for-fee QoS is the best way to give startups the wherewithal to compete against Google’s unregulated private network. I wonder why more people don’t see that.

The concern I have has to do with prioritization of traffic based on the creator of the content or service. It makes sense that network operators might decide to deprioritize asynchronous traffic (email, usenet) to allow syncronous traffic (voip, streaming video) to get through… or that an overloaded network might decide to degrade high-bandwidth, syncronous traffic to lessen network load.

What worries me is situations like the one I referred to in the Inc. editorial where Shaw may be disguising business decisions as technical decisions – deprioritizing VOIP traffic from systems like Skype or Vonage in order to better market their proprietary VOIP services. This really complicates the sorts of decisions a user has to make in choosing an ISP – if she’s trying to figure out which applications work best on which network, as well as pricing and service, it becomes a very difficult proposition. And it raises barriers of entry for content creators if getting content delivered on a particular network requires a business relationship with that network operator.

My reason for concern comes back to my African experiences. ISPs wanted to offer VOIP as an affordable way to make overseas calls. Monopoly telcos – using QoS as a rationale – tried to block VOIP… and succeded in some cases. This was a major loss for consumers – competition from legal VOIP would have brought down telco costs for all consumers. (Eventually, it did – huge numbers of people used VOIP even though it was illegal…)

My interest in network neutrality has to do with ensuring that it’s easy for people to create new services on the net with a minimum of barriers. I agree that there’s quite a bit of hype getting tossed around on both sides of this debate. I’m not sure terming me a “neutrino” and suggesting that I’m being intentionally deceptive is helping the debate, Richard.

As Lippard says, you clearly don’t understand how the Internet works today. Your claim: “Right now, no major ISPs give preferential treatment to one type of traffic over another” is false at several different levels. The cable company access lines carry voice, video, and packet traffic on different channels and with different QoS. The DOCSIS protocol that carries packet traffic on this network is isochronous, and it most certainly prioritizes based on QoS. The telco lines that carry DSL carry analog voice in a different FDM domain, and apply completely different rules to it. And it’s always been legal for any ISP to prioritize traffic any way he sees fit; common carrier rules didn’t address ISP filtering and forwarding. WiFi networks also prioritize based on QoS requirements through a little feature called 802.11e.

VoIP doesn’t inherently work well on a packet network, with first-come, first-served queuing when load is reasonably high. Providing priority queuing is the most efficient way to combine VoIP with the heavy demands for bandwidth that HDTV poses. Simply throwing bandwidth at this problem doesn’t solve it, and when the bandwidth has to be bought with somebody else’s dime, your solution is a non-starter.

Every net neutrality argument I’ve seen relies on spin and misinformation, and yours is no exception.

Not true–there are no less than six bills in Congress to mandate various forms of “net neutrality” through new regulations.

“Right now, no major ISPs give preferential treatment to one type of traffic over another. The proposals to do so are why the net neutrality folks are getting worried.”

Again, not true in the backbone world–Global Crossing, Level 3, AT&T, MCI, Qwest, and others offer products that use QoS and separate MPLS VPNs for certain types of traffic (and permit customers to do the same with their own traffic). At least one of the above-mentioned bills (HR5417) will classify backbone providers as broadband providers and make these products illegal or of questionable legality, which makes no sense.

I manage security for a global backbone (Global Crossing). If we didn’t keep IP-VPN, IP-Video, and VOIP distinct from public Internet traffic, those services would not be as reliable. The public Internet contains huge amounts of malicious traffic–there are tens of millions of compromised consumer devices connected to it, being used in botnets to send spam, phishing attacks, and launch denial of service attacks. By keeping those services distinct, those attacks don’t impair services that are very sensitive to latency and jitter like real-time voice and video.

Any net neutrality bill that bans the use of QoS as backbones are already using it is a bad bill.

The history of the carriers’ behavior is not proud. Columbia’s Tim Wu points out that before net neutrality was enunciated as an FCC principle and enforced:

– AT&T warned customers that using Wi-Fi home networking equipment was a ‘federal crime’
- Cox Cable disciplined users of virtual private networks
- Comcast blocked Internet VPN ports, which prevented Washington state workers from telecommuting

Now revisit the COPE Act and try to find the following words:

– block
– impair
– degrade

If the carriers don’t intend to discriminate (block, filter, degrade, or impair various content providers), why have they wall-papered Washington with green and, coincidentally, had those very words omitted from the COPE Act?

Other questions that consumers need to ask themselves:

– when the carriers — and not the consumers — get to discriminate between content-providers, what happens to ventured-funded innovation, startups, and America’s technological leadership position?

– does it make sense that the telcos are trying to shoehorn TCP/IP into an HDTV/IPTV delivery platform when they haven’t even mastered QoS for VoIP on private, dedicated networks?

The sad part about all of this is that the companies owning the larger capacity ATM networks on which the backbone of the internet is built and the bulk of content is hosted will not even have to degrade VoIP service to kill smaller companies like Skype/Vonage. The cost of the reliable low-latency access to the backbone that they need will kill them under the proposed model long before people can complain about call quality. If they cannot compete on price, their only advantage at this point, these services will be dead shortly. I also find the false characterization of the issue as a life and death one pretty funny. My take is that any medical monitoring company that depends on the “public” internet – for communications deserves whats coming to it when the “free” lunch runs out.

What I liked best is the fact that you latched on to the VoIP in Africa aspect. VoIP promises to put sub-saharan African international telecom pricing within much easier reach of the common man. There is absolutly no doubt in my mind that if things are left as they are, there will be a mini-revolution around VoIP just as there was and continues to be a mini-revolution around cellular telephony.

The most painful aspect of all of this from a developing world perspective is that just as the cost of internet access is slowly but surely falling to decent levels thanks to affordable last mile solutions like hybrid & tuned WiFi, WiMax, ADSL & HDSPA coupled with “transport” over cheaper high-capacity VSAT access and capacity on existing undersea cables (and of course, the promise that is EASSy), these prices will once again rise to astronomical levels thanks to increased backbone access charges under the proposed “pay-for-what-you-eat” model.

Steve

Actually, keeping things the way they are now is precisely what most of the net neutrality people are advocating. Right now, no major ISPs give preferential treatment to one type of traffic over another. The proposals to do so are why the net neutrality folks are getting worried.

As for the “separate access lane” – this gets proposed roughly every five years in the history of the Internet. Ten years ago, you could read learned white papers on why the public internet would never support voice traffic and why ATM networks with prioritization were neccesary for that sort of traffic. A few years back, you could read arguments that firesharing traffic would destroy the net as we know it. The truth is, the net has proven surprisingly resilient and able to adapt to new uses, so long as no one writes legislation prohibiting or making unlikely those uses.

As for the separate access lane for medical data – this just doesn’t make technical sense. The sort of data we’re talking about are very small packets of info sent pretty infrequently – once per second, for instance. A network that can handle video – dozens of packets per second, large amounts of data – can handle this sort of data without reconfiguration unless the whole network is saturated to the point of unusability. I really think it’s a red herring, not a legit argument.

Meanwhile, I still fail to see what’s so scary about keeping things the way they are now. Especially when legislating net neutrality means no separate access lane that could be used by hospitals to monitor patients from afar, without fear of being interrupted by P2P/BitTorrent downloading.

If Google etc. aren’t “big bad” then neither are Verizon etc.