The expressive power of the Internet design

The present Internet is not defined in terms of its semantics, at least at the packet level. The loose packet carriage model of “what comes out is what went in” is intentionally almost semantics-free. The packets just carry bytes. Packet boundaries can have some limited semantics, but not much. The...

Full description

Bibliographic Details
Main Author: Clark, David D.
Format: Working Paper
Language:en_US
Published: © Massachusetts Institute of Technology 2022
Online Access:https://hdl.handle.net/1721.1/141690
_version_ 1811090586732593152
author Clark, David D.
author_facet Clark, David D.
author_sort Clark, David D.
collection MIT
description The present Internet is not defined in terms of its semantics, at least at the packet level. The loose packet carriage model of “what comes out is what went in” is intentionally almost semantics-free. The packets just carry bytes. Packet boundaries can have some limited semantics, but not much. The original design presumed some constraints on the semantics of packet headers, such as global addresses, but the progress of time has violated these and the Internet keeps working. TCP does impose some modest semantic constraints, but of course TCP is optional, and not a mandatory part of the architecture. What defines the Internet, and the range of behavior that is available in the Internet, is the expressive power of the packet header, which has more to do with its format than any semantics. Most fields (e.g. packet length) are unremarkable, some (like the TOS bits) have been redefined several times in the history of the Internet, some (like the options) have atrophied, and some (most obviously the IP addresses) have had a most interesting history in which the only constants are that they are 32 bit fields, that whatever value they have at each end must remain constant for the life of a TCP connection (because of the pseudo-header) and that at any locale in the network, they must provide the basis for some router action. They can be rewritten (as in NAT), turned into logical addresses (as in multicast or anycast), and they can be microcoded in a number of ways to capture address hierarchy (net/rest, A/B/C, CIDR). All that really matters is that they are 32 bits long, and that at any point, they must have at least local meaning to a forwarding process.
first_indexed 2024-09-23T14:48:33Z
format Working Paper
id mit-1721.1/141690
institution Massachusetts Institute of Technology
language en_US
last_indexed 2024-09-23T14:48:33Z
publishDate 2022
publisher © Massachusetts Institute of Technology
record_format dspace
spelling mit-1721.1/1416902022-04-07T03:11:55Z The expressive power of the Internet design Clark, David D. The present Internet is not defined in terms of its semantics, at least at the packet level. The loose packet carriage model of “what comes out is what went in” is intentionally almost semantics-free. The packets just carry bytes. Packet boundaries can have some limited semantics, but not much. The original design presumed some constraints on the semantics of packet headers, such as global addresses, but the progress of time has violated these and the Internet keeps working. TCP does impose some modest semantic constraints, but of course TCP is optional, and not a mandatory part of the architecture. What defines the Internet, and the range of behavior that is available in the Internet, is the expressive power of the packet header, which has more to do with its format than any semantics. Most fields (e.g. packet length) are unremarkable, some (like the TOS bits) have been redefined several times in the history of the Internet, some (like the options) have atrophied, and some (most obviously the IP addresses) have had a most interesting history in which the only constants are that they are 32 bit fields, that whatever value they have at each end must remain constant for the life of a TCP connection (because of the pseudo-header) and that at any locale in the network, they must provide the basis for some router action. They can be rewritten (as in NAT), turned into logical addresses (as in multicast or anycast), and they can be microcoded in a number of ways to capture address hierarchy (net/rest, A/B/C, CIDR). All that really matters is that they are 32 bits long, and that at any point, they must have at least local meaning to a forwarding process. This material is based on work supported by the U.S. Office of Naval Research, Grant No. N00014-09-1-0597. Any opinions, findings, conclusions or recommendations therein are those of the author(s) and do not necessarily reflect the views of the Office of Naval Research. 2022-04-06T11:12:23Z 2022-04-06T11:12:23Z 2009-04-27 Working Paper https://hdl.handle.net/1721.1/141690 Clark, D. D. (2009). The expressive power of the Internet design (ECIR Working Paper No. 2009-1). MIT Political Science Department. en_US ECIR Working Paper No. 2009-1 Attribution-NonCommercial-NoDerivs 3.0 United States http://creativecommons.org/licenses/by-nc-nd/3.0/us/ application/pdf © Massachusetts Institute of Technology
spellingShingle Clark, David D.
The expressive power of the Internet design
title The expressive power of the Internet design
title_full The expressive power of the Internet design
title_fullStr The expressive power of the Internet design
title_full_unstemmed The expressive power of the Internet design
title_short The expressive power of the Internet design
title_sort expressive power of the internet design
url https://hdl.handle.net/1721.1/141690
work_keys_str_mv AT clarkdavidd theexpressivepoweroftheinternetdesign
AT clarkdavidd expressivepoweroftheinternetdesign