Sunday, January 2, 2011

IPv6: Reloaded

After looking at the advocacy and reasons given even the incompatibility factor, I smell something fishy with this IPv6 addressing system implementation.

The push does not seem to come from an engineer. And engineer would take seriously the issue of compatibility especially a computer engineer.

It is a core part of his training for an engineer to design a change over plan that minimises costs and ensures smooth transition from an old system.

It is a grave engineering mistake to disregard compatibility. So many things would break resulting in great losses, both financial and with high possibility, life.

However, politicians can demand such a push and in their political mind would prefer mass media to do drive the change over than practicality of the solution.

It raises questions like how much control can one have over an IPv6 network compared to IPv4? What would the reason for jumping IPv5 when it seems most problems IPv6 is solving can be solved by the combination of IPv4 and IPv5?

We might fall into a trap we understand not.

I thought about the internet. Why do we need it? Communication? True. With who? I found that my average communication I do on the internet is still with my fellow malawians. But the tools I communicate with are not in Malawi but outside. Yes, they are free tools but the fact that they are free, is the reason I use them. The fact that I use them, makes me need to connect outside Malawi which at the end of the day, at national level, it is a great expense.

The allocation of addresses Malawi has for the number of ISPs available (those are who need public up addresses) I feel should be enough. We have less than 100 ISPs. How many ip addresses is that? If we might need more than 500 public ip addresses, the somewhere somehow is a big abuse of ip addresses. Something MACRA might want to look at. A proper design of the domain name system would enable us to reduce dependence on the outside network and therefore reduce the cost of connectivity. Most of us are forced to use the freely available tools because few good are provided and in a restrictive manner which with reduced dependence on outside internet would be available cheaper. There have been talks of a national internet exchange provider which allows for local exchange of local content. We have had a national infrastructure(MTL) which against all economic reasons was privatized. It is good that MACRA is creating a fibre network but much has been invested in creating the previous infrastructure and for a long time making it very possible MTL was for ill reasons privatized and at very small value than the infrastructure available.

It seems even companies like Airtel don't value infrastructure much than they do a customer base. They can overhaul the whole network(it is easy anyway you just buy new access points and connect them to talk to each other over already built transmitters). How much would an access point cost? Not worth the worry I think.

So what is the cost of building a transmitter for some radius? What is the cost maintaining it? What would be the cost of using it? Take away the cost of connecting outside Malawi, the cost communication is unbelievably low. Put in place the tools for communication locally, dependence for external tools is little. Now having reduced the dependence of external tools, the need for external communication is further reduced and by a great percentage I think. When external communication outlets are reduced, the cost of maintaining such shared on over the whole population of Malawi. This by a great percentage would bring the cost of communication. That is what fuels economic development.

I wish government, would allow me even fund me to implement such for the population of Malawi. Not single handedly but I can get a work force to achieve this. We will buy same tools as other networks and maybe cheaper and better, but we will manage them better and reduce the cost greatly.

My contacts are down below in one of the posts.