Primary Domain
Internet Infrastructure
Within the Primary Domain facet, Internet Infrastructure intelligence groups reporting by primary domain so readers can follow a focused area of internet infrastructure, governance, connectivity markets, or digital capital. The page brings together related articles, public evidence, institutions, companies, people, regional exposure, operating dependencies, and market context that may otherwise sit across separate category pages. It explains the domain, the likely actor class, the market or governance context, and the source material readers should use when comparing signals. Operators, analysts, and governance readers can see how the same domain appears across events, profiles, market shifts, public-source evidence, regional dependencies, and longer-cycle infrastructure decisions over time.
History
The Address That Asked the Wire: How ARP Made Local Reachability Revisable
Routing can choose the next IP address without knowing how to put a frame on the local cable. ARP filled that last gap with a public question, a private answer, and a memory designed to be corrected.
History
The Connection That Had to Prove It Was New: Why TCP Needed a Three-Way Handshake
A packet can arrive at the right address, carry a plausible number and still belong to a conversation that has already died. TCP's familiar opening exchange was designed around that ambiguity: before accepting a new connection, both endpoints must show that they have seen each…
History
The Packet That Had to Learn Its Size: How Path MTU Discovery Moved Fragmentation to the Edge
A packet may cross a path whose narrowest link is invisible to its sender. The Internet spent four decades changing who should discover that limit, who should report it and who should decide what size to send next.
History
Two Kinds of Nothing: How DNS Learned to Separate NXDOMAIN from NODATA
An empty DNS answer can conceal two opposite realities: the requested name may be absent altogether, or it may be alive and carrying every record except the one requested. The Internet had to make that difference portable, cacheable and temporary before saving queries could stop…
History
The Error That Learned to Expire: How DNS Cached What Did Not Exist
The expensive DNS question is not always the one with a large answer. It may be the miss that thousands of resolvers repeat because nothing can be stored. Negative caching gave absence a scope, evidence and an expiry time—then forced operators to accept that a newly created name…
History
The Core That Had to Stop Being the Core: EGP and the Invention of Autonomous Systems
The early Internet did not decentralize by pretending that every network agreed. It learned to survive disagreement. The Exterior Gateway Protocol drew a boundary around what one routing administration could keep private and what it had to tell its neighbours. That boundary made…
History
The Mail Address That Outlived Its Server: How MX Records Separated Identity from Delivery
An email address looks like a destination. In the early Internet, it often was: a mailer could take the domain after `@`, open SMTP to the machine with that name and hope the mailbox lived there. In January 1986, RFC 974 described why that assumption had stopped scaling—and how…
History
The Clock Built from Disagreement: How NTP Made Internet Time Without Trusting One Server
In 1989, an Internet-wide experiment found clocks wrong by minutes, hours and, in a few cases, weeks. The problem was not a shortage of machines willing to state the time. It was deciding which statements deserved to move a local clock when packets, oscillators and operators…
History
The Web CERN Chose Not to Own: How a Laboratory Designed Its Own Exit
The World Wide Web did not become universal merely because it was invented at a famous laboratory. It became portable across institutions because its common machinery was kept small, its specifications were public, its clients escaped one workstation, and CERN eventually removed…
History
The Network That Let Institutions Join Unequally: CSNET and the Architecture of Partial Access
CSNET widened the networked research community without pretending that every institution could afford or operate the same connection. By combining ARPANET access, full Internet service over commercial X.25, and dial-up PhoneNet mail, it showed that interoperability could begin…
History
The Map Made from Expired Packets: How Traceroute Turned Failure into Internet Evidence
In December 1988, Van Jacobson was not trying to draw the Internet. He was trying to find out where his packets had gone. The tool he assembled at Lawrence Berkeley Laboratory made the network answer indirectly: send a packet designed to expire, listen for the complaint, then let…
History
The Day the Old Protocol Lost Its Network: The 1983 TCP/IP Cutover and the Limits of Technical Authority
On 1 January 1983, an ARPANET host that still depended on NCP faced a new fact: the network beneath it could stop servicing the old protocol. The famous cutover mattered not because a memo became universal law, but because a disclosed rule reached an execution boundary whose…
History
Where the Network Stopped: The IMP, the Host, and the First Price of Shared Correctness
ARPANET’s Interface Message Processor made packet switching operational by putting substantial machinery into a dedicated subnet. It also made a harder problem visible: infrastructure can carry, check, meter and protect communication without possessing the knowledge required to…
History
When One Router Announced Everyone’s Network: AS7007 and the Failure of Unchecked Trust
When One Router Announced Everyone’s Network: AS7007 and the Failure of Unchecked Trust intelligence summary explains the development, the public evidence available to readers, the organisations involved, the regional context, market exposure, and the infrastructure consequences…
History
When the First Network Became Replaceable: ARPANET’s Retirement and the Internet That Continued
ARPANET did not take the Internet with it when it was retired in 1990. That was not because the pioneering network had become unimportant. It was because its most important function—making unlike networks communicate—had already escaped the machinery and institution that first…
History
The Window That Learned Restraint: How TCP Stopped Congestion Collapse Without a Traffic Authority
The Internet's answer to its first congestion crisis was not a controller above the network. It was a new discipline inside every sender: learn the path from acknowledgements and loss, increase carefully, and retreat before retransmissions turn shared capacity into useless…
Asia-Pacific Cloud Services Trends
Cloudflare's Hong Kong incident exposed a gap between location and state health
Cloudflare restored “Durable Objects” in Hong Kong in about 21 minutes. The sharper lesson sits in the status record: the stateful product was degraded while the Hong Kong location and the named Realtime components remained green.
Creators
Eric Dumazet: Linux’s hidden queues
TCP can make a reasonable sending decision and still lose time after packets leave its direct control. Eric Dumazet’s work on TCP Small Queues, fair flow scheduling, pacing and cache-aware data structures changed how Linux accounts for this hidden packet backlog, although user…
Creators
Sage Weil: the engineer who made distributed storage computable
Sage Weil’s most enduring contribution was not a storage device but an architecture. CRUSH calculates data placement, while RADOS distributes storage, repair and movement across the cluster. Ceph enabled ordinary machines to operate as one system, then had to learn to continue…
Creators
Sage Weil: The Architect Who Made Distributed Storage Calculable
Sage Weil’s most durable contribution to internet infrastructure was not a storage appliance but an architecture. Ceph used CRUSH to calculate where data belongs and RADOS to spread storage, repair and movement across a cluster. The design helped commodity machines behave like…