Primary Domain
Technology
Within the Primary Domain facet, Technology 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.
IETF
Jari Arkko and the Registry Field That the Protocol Did Not Need
A protocol registry is often read as a durable record, which makes every old column look like a permanent requirement. RFC 8602 takes the opposite discipline: an inherited postal-address field in two TRIP registration procedures had no identified use, so IANA stopped collecting…
IETF
Lars Eggert and the UDP Datagram That Did Not Escape Shared Costs
UDP gives an application a small transport contract. It does not give it a private network. Lars Eggert's work on RFC 8085 makes that distinction operational: the freedom to choose datagrams leaves congestion responsibility with the traffic that actually shares a path.
IETF
Nandita Dukkipati and the TCP Recovery That Stopped Lurching
A sender can finish loss recovery at exactly the intended congestion window and still have taken the wrong journey there. A long pause followed by a burst and a sequence of small, ACK-clocked releases may move the same number of bytes. They do not place the same strain on a…
IETF
Ashesh Mishra and the BFD Check That Had to Be Reversible
The receiver had not accepted the packet. It was merely calculating whether it should. Yet that calculation could erase the only state from which a rejection remained possible.
IETF
Wes Hardaker and the DNS Server That Had to Outlive Two TTLs
The new nameserver was answering perfectly. The old one was quiet. An operator shut it down after the child zone's TTL elapsed—and a minority of users vanished, because their resolvers were still keeping time from the parent.
IETF
Mirja Kühlewind and the QUIC Spin Bit That Measured the Application Period, Not Network RTT
A passive observer saw a clean pulse every 200 milliseconds and labelled the path RTT 200 ms. The packets were real and the arithmetic was correct. The conclusion was not: a sparse application was sending every 200 ms across a much faster path.
IETF
Murray Kucherawy and the DKIM Signature That Left the Tail Unsigned
The verifier returned `dkim=pass`. The message was therefore signed — but not necessarily all of it. One optional number can make both statements true: `l=` tells DKIM to stop hashing after a specified number of canonicalized body octets, while the reader may still see everything…
IETF
John Klensin and the SMTP Reply That Accepted Responsibility, Not Delivery
The sending server sees `250 OK`, removes a message from its queue and calls the transaction complete. That can be correct. Calling the message delivered is not. In SMTP, the final positive reply moves custody to a new operator; the destination still has to be reached, and…
IETF
Tomek Mrugalski and the DHCPv6 Success That Did Not Renew the Lease
A laptop changes networks and asks whether its existing IPv6 addresses still belong on the new link. A server answers `Success`. The tempting conclusion is that the leases are safe. RFC 9915 says something narrower: the addresses fit the link, while every old lifetime keeps…
IETF
Bob Briscoe and the L4S Mark That Did Not Prove Low Latency
An IP packet can carry the L4S identifier all the way across a network even when one operator deliberately sends it through the Classic queue. That is not a contradiction. It is the clue that a packet's declared treatment, its actual queue and its measured delay are three…
IETF
Kent Watsen and the UDP Model That Did Not Record the Running Socket
A management system can hold a perfectly valid UDP endpoint configuration while the machine is listening somewhere else—or nowhere at all. RFC 9984 makes the common vocabulary portable, but it deliberately leaves the runtime testimony to the system that executes it.
IETF
David Schinazi and the UDP Tunnel That Opened Before the Destination Answered
A green CONNECT-UDP response can arrive before the destination has sent a single packet. RFC 9298 makes that apparent paradox precise: the proxy is reporting its own readiness, not certifying a remote application's existence.
IETF
Christopher A. Wood and the Privacy Boundary No One Operator Should Hold
Oblivious HTTP does not ask an operator to forget what it can see. It rearranges the transaction so that one operator can see where a request came from and another can see what the request says—then makes the absence of a single complete view the privacy control.
IETF
Martin Thomson and the Key Log That Could Decrypt a Session but Not Prove It
The incident package looked decisive: a packet capture, a small text file and a screenshot of readable TLS traffic. The decryption had worked. Yet the file did not say who enabled logging, which endpoint produced the secret, when it was collected or whether the alleged session…
IETF
Todd Herr and the DMARC Pass That Did Not Make the Message Safe
The green word `pass` can be perfectly correct and dangerously misunderstood. In DMARC it certifies a narrow relationship between domains. It does not certify the person in the display name, the proposition in the message, the attachment behind the button or the wisdom of putting…
IETF
Adrian Farrel and the Implementation Receipt That Disappeared Before Publication
One of the most useful sections in an Internet-Draft is written with an expiry condition. It can name the code, version, coverage, licence, test contact and interoperability evidence that made a proposal real. Then, if the proposal becomes an RFC, the section is supposed to…
IETF
Lukasz Kondrad and the RTP Group That Was Not Yet a Reconstructed Scene
A session description can place an atlas, occupancy, geometry and colour stream inside one V3C group. That is a precise statement about membership—not evidence that a receiver rebuilt the same three-dimensional scene.
Global Institutional Trends
OpenAI coalition urges faster AI cyber defence
More than 100 organisations are calling for faster vulnerability repair, shared threat intelligence and wider access to defensive AI for critical infrastructure.
Asia-Pacific Cloud Services
Velo Technologies expands its infrastructure business from Cyberjaya
The Malaysian private company spans data-centre capacity, cloud, managed security, networks and IT outsourcing. Its partnerships are increasingly visible, but its ownership names, absolute financial scale and the economics behind some of its infrastructure commitments remain…
North America Cloud Services
FreeCast targets telecom operators with streaming PaaS
FreeCast is pitching its streaming platform to telecom and broadband providers, with Assist Wireless and enTouch Wireless already signed to use the service.