smackfoo.com | I find your lack of faith disturbing.

I’m a recent convert and daily reader of Technology Evangelist. Now, going by recent posts you may think that odd given some of my responses — but I respect the viewpoints expressed and the through that goes behind many of the posts.

However, I do not always agree with the summations made — the recent thoughts on prioritized bandwidth entry is a case in point.

As is always the way — a few misconceptions made at the emergence of the “network neutrality” argument have now become so ingrained they’re actually warping what is really at stake. Lets take a moment to run through the crux of Ed Kohler’s thoughts and how they stack up against the existing inkernet we all know and share a love/hate relationship with. There are a number of points Ed makes, but I’m going to focus on the two biggest misnomers of the increasingly confused stance that is “ network neutrality“.

• Multiple Classes of Service: Telecommunications companies suggest that networks should adopt models similar to airlines, with business and economy classes. Telecommunications companies are in more trouble than I thought is they’re looking to the airline industry for business models.

Firstly, this is a bit of a furphy and misnomer. QoS is the equivalent of an airline Business Class and Coach with a few extra levels or dynamic seating arrangement in between, for good measure. Quality Of Service is a technology that has existed for many years and is an integral part of the internet topography. It ensures that Voice over IP, Video on Demand, or IPTV packets have the uber low latency (occasionaly refered to as ‘ jitter‘) needed to work reliably, yet still allow the torrent and porn industry to flourish — said data is thus frequently prioritised based on it’s latency requirement. Data transfer and general web surfing does not require an overly robust or highly efficient network - indeed HTTP type traffic emerged during the times of ISDN, packet switching and analogue modems. It’s robust enough to work, virtually no matter how bad the latency between two points is.

The idea that this is a fundamentally flawed concept is ridiculous given it’s the very thing that allows you, right now, to check your email, surf the web, download some data and perhaps watch the fairer sex in it’s native habitat doing.. something.. ooh err.. is that even possible?! Err, right - yes, as I was saying.. It’s the very thing that allows you to do several things at once and compartmentalisation of data into “virtual” circuits for VoIP, IPTV that is seperated from generic web surfing traffic, for example is actually a very positive step.

It allows latency sensitive applications to function perfectly, without torrent, porn, email or data transfers impacting on such latency sensitive technologies. It all works seamlessly in the background. With ISPs and third party companies all beginning to push heavily into link sapping technologies such as IPTV — getting that data into a more manageable structure is a very progressive move.

This has nothing to do with, nor should ever be confused with a premium versus pauper network — again such networks do exist and have been around for some time. There are numerous businesses that provide high grade networks to ISPs, Corporates and other industry players. They pay a premium for a premium network. Again this has been a common part of the internet and WAN technologies for many years. Is a high-performance McClaren F1 the same cost as a second hand Lada? No - yet that is exactly what network neutrality proponents are upset about. Lets leave the technology and topography that’s been in use for years out of the argument — it’s not relevant to the real cause of concern.

The real reason for the push for premium/pay-per-use (not quality based) type networks is simple. Cost. Our rampant use of bandwidth sapping technologies has lead to the inevitable result many could see coming — expansive types of data are being carried over standard internet links which can cost a small fortune to maintain. Routing a lot of data like VoIP, IPTV, VoD, etc on separate dedicated or virtual networks where data costs can be accurately measured (which leads to cost reductions) is a very effective method of rejuvenating existing network links. It reduces load on core internet nodes, while shifting voice and video into a more suitable conduit.

Again, that’s not something new. Video has been carried over cable networks for years, whilst voice has run over various PSTN networks for even longer — completely seperated and isolated from internet traffic.

Cost has driven companies to look at premium/pauper networks to try and regain some control over their networks. Embracing a more open, virtualised network, where bandwidth is split across multiple technologies, rather than lumped togther with QoS trying to keep control, is a far better option - so lets drop the FUD from the argument and stick to what works.

• Mission Critical Applications: Mark Cuban should take a look at how mission critical services are delivered offline. We don’t build separate roads for ambulances. We build an infrastructure that’s capable of carrying everything from a groups of teenagers to movies, to ambulances racing to save a lives. Is it perfect? No. But it’s much more efficient and cheaper than building dedicated lanes for mission critical services like emergency vehicles.

Actually - I think you’ll find Mark isn’t just whistling out his proverbial. Many countries use helicopters and aircraft for lightning response in critical emergencies. Are you suggesting that we should instead just send ambulances no matter the terrain or situation? How about we all use exactly the same model vehicle for all ambulances, that operate at the same range, with the same speed and see how many people die because we get hung up on all the wrong ideals.

Again - the misnomer of a “wholly separate network” has become confused and mixed up with quality based networks. We need to apply some form of “future thinking” to the internet landscape - that means intelligent handling of data. At some point QoS and related technologies will no longer be sufficient. We need something that can expand, we need redundancy and we need to keep the clock ticking. If the internet stops, so does our civilisation that is increasingly bound to it. If we spent more time embracing an open, varied topology that separates data into more logical flows, rather than panicking about a future internet that may restrict what we can do, this whole issue would be a non-event.

Net neutrality is in essence “the internet remaining all things to all people”. For that ideal to grow, it needs a topology and topography that separates data into logical networks, virtual or otherwise, it needs funding and it needs to cope with tomorrows world of on demand everything.

Lets stop the FUD, lets stop debating semantics and side issues that have been confused from the start and start pushing for a global network that logically handles our needs now and into the future.