So says a Study on the Internet’s Technical Success Factors commissioned by APNIC and LACNIC – the regional internet address registries for the Asia–Pacific and Latin America and Caribbean regions respectively – and written by consultancy Analysys Mason.
The document states that “a significant fraction of global IP traffic now consists of data that is moved between the datacentres and edge networks of large internet companies.” Those companies’ needs, and growing networks, lead the analysts to suggest that "over time, we could see the internet transform into a more centralised system with a few global private networks carrying most of the content and services.
Another risk is that when private networks break, many users suffer. Exhibit A: yesterday’s AWS brownout, which hurt Netflix and Disney+, among others.
Yet, if you look at nearly all the alternative social networks springing up, you’ll see decentralisation, openness, interoperability, chronological feeds, no Big Tech…
See https://www.theregister.com/2021/12/09/study_on_the_internets_technical_success_factors/
#technology #interoperability #BigTech #openstandards #decentralisation
If BigTech will continue to centralize and ignore the basic concepts that made the internet possible their internet will be down way more often. They are actually shooting themselves in their foots in slow motion :D The issue I have is that our daily lives depend more and more on this corporate networks. A responsible state would not put all their eggs in that basket and stick with the “old” internet.
"In this scenario, what remains outside these private networks are primarily ISP networks that move traffic to and from end users, and the user experience would be shaped by how close a user sits to the private network of the relevant internet company.”
This is already a thing. All major CNDs bind with major ISP directly via major internet exchanges. Netflix and Google put their content servers (big caches) directly in the ISP networks already; otherwise the internet backbone would not be able to handle all the traffic long time ago.
That may be so, but the outages also have a way bigger impact. The question is where the exclusivity will lead us… to privately owned routing (for those who can afford it)?
Good question. I hope not. For now corporations are shaping the routing but I think we are not fully dependent on them yet. But it feels like the forces are stacked against the principles that made the internet and if we don’t protect them the outcomes will be very bad.
Yes it does appear that whether governments or corporations, they tend to have the loudest voices and most funding…
It’s similar to the supply chain issues. Centralization and concentration are good for efficiency but they can lead to very fragile systems. When those fragile and now big systems break, they make for high profile events that serve as a warning and counterforce to efficiency.
It doesn’t matter if you shave 0.1% off your bandwidth costs if you lose 5% of your customers every year due to outages.
So I think it will work itself out as long as people cancel and look for alternatives these things when they’re not getting a good experience. Normalizing consumer rights is the best way to help this and plenty of other issues.
So looks like we have the internet represented by *NIC etc. that is political. The one that is build to top that is corporate… and some “blockchain” based on on top of that further that will be owned by cryptobros… got it.
Are regional internet address registries political?
Yeah, what I mean is that the “control” over the internet is based on participation of organizations controlled by each individual state (hopefully represented by democratically elected government, representing people etc.) in contrast to handful of global corporations that has escaped such governance. This makes the internet slow to move forward (e.g. IPv6, DNSSEC etc.) in contrast to “move fast and break things” (e.g. the CDN overlay, QUIC, HTTP/3 protocol etc.). So there is completely different dynamics, control and outcomes for us.
