[ih] History of the Economics of the Internet?

[John R. Levine]

On Fri, 31 Jul 2026, Brian E Carpenter wrote:
>> My impression is that for Internet interconnections, if the two parties are 
>> of comparable size they peer, or if one is bigger, the smaller pays the 
>> larger for transit.  I've also heard of paid peering, somewhere in between 
>> but not common.
>
> And they are always confidential bilateral agreements, not multilateral,
> because otherwise they would be an illegal cartel.

My impression is that it's common knowledge that at some IXes it's all 
peering but yes, they're not public.  I have also heard that a surprising 
number of peering agreements are just a handshake, no contract.  If no 
money's changing hands, there's not much to contract about.  If the other 
side does something nasty you can just disconnect.

> Also, "peering" and "transit" are economic terms, not technical ones.
> "BGP peering" is a technical term, applicable to both economic peering and
> transit.

Right.  If you're big enough to peer you're probably big enough to be 
defaultless, while a lot of transit just sends everything upstream, but 
there's no necessary connection between the business model and the network 
setup.

R's,
John

> On 31-Jul-26 11:49, John R. Levine via Internet-history wrote:
>> On Thu, 30 Jul 2026, Scott O. Bradner wrote:
>>> it also touches on what the legacy telcos expect as to settlements etc
>> 
>> Right, but that's telephone call settlements, not Internet data
>> settlements.  Ineffiecnet little telcos like mine are sad not to get
>> settlements.  I hear there are (maybe were) tiny phone co-ops in the
>> midwest where not only is your phone line free, you get regular dividends
>> paid from the settlements.
>> 
>> My impression is that for Internet interconnections, if the two parties
>> are of comparable size they peer, or if one is bigger, the smaller pays
>> the larger for transit.  I've also heard of paid peering, somewhere in
>> between but not common.
>> 
>> R's,
>> John
>> 
>>>> On Jul 30, 2026, at 4:48 PM, John Levine <johnl at iecc.com> wrote:
>>>> 
>>>> It appears that Scott Bradner via Internet-history <sob at sobco.com> said:
>>>>> this covers some of the current discussions over the Internet's economic 
>>>>> model
>>>>> within the regulatory world (at least in the US)
>>>>> 
>>>>> https://isoclive.substack.com/p/fcc-ip-transition
>>>> 
>>>> Not really. It's about moving the telephone system from legacy SS7 to IP
>>>> signalling, switching, and transport. It's the same technology as the 
>>>> Internet
>>>> but it's separate dedicated IP networks with privately negotiated
>>>> interconnections and occasionally still per-minute settlements.
>>>> 
>>>> There is also telephony that runs "over the top" of the public Internet, 
>>>> but
>>>> that's something else. That's VoIP, Voice over IP, while the former is 
>>>> VuIP,
>>>> Voice using IP. There are plenty of vendors that provide VoIP to VuIP 
>>>> gateways
>>>> so your VoIP terminals can make and receive calls to or from real phone 
>>>> numbers.
>

Regards,
John Levine, johnl at taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly