Profile
Briefing
Briefing profile intelligence connects person, company, institution, and other entity coverage for readers comparing the subjects behind BTW.MEDIA reporting. The page helps compare public profiles, related events, evidence packs, role changes, operating footprints, governance exposure, organisational links, and market relevance while keeping organisation and people records distinct from ordinary news chronology. Readers can understand the subject behind the headline, the evidence supporting that profile, and the infrastructure or market relationships that make the profile relevant. They can see who or what the profile represents, which public sources support it, how it connects to companies, markets, governance bodies, or events, and what changes may deserve continued attention as the subject evolves.
North America Datacenter Trends
AIB’s $1.32bn Nebius contract still needs project finance
The binding 50 MW agreement changes AIB’s customer status, not the funding status of its planned data-centre build: the company’s own illustrative financing model leaves most project capital unarranged.
North America Datacenter Trends
Applied Digital’s 250 MW milestone puts rent conversion in focus
Applied Digital says another 75 MW at Polaris Forge 1 is ready for service, lifting reported operational load to 250 MW. The milestone advances a 400 MW contracted campus, but the next test is what the new halls add to base rent—not a forecast built from megawatts or fit-out…
Africa National Telecom Trends
Vodacom’s MIMO trial answers a space problem, not the rollout economics
Vodacom says a compact passive-active antenna tested with Nokia in the Western Cape can add capacity at constrained sites, using 2.6 GHz and 3.5 GHz. It has not disclosed the site count, capacity lift, timetable or cost; the announcement establishes a deployment option, not a…
IETF
The Network Map Now Needs a Traffic Budget
A live infrastructure map is useful only if it can keep up with change. A new requirement in an IETF SIMAP draft asks what happens when the service that supplies that map becomes a source of network load itself.
IETF
TCP-AO's KMAC Name Was Not a Complete Recipe
An algorithm label can look precise while leaving a key derivation ambiguous. The latest TCP-AO working-group draft names the NIST KMAC variant behind its proposed 128-bit authenticator, a small textual change with a concrete interoperability question for implementers.
IETF
The RadSec Reply That Must Die Without Killing the Link
A late RADIUS response can fail an authenticator check because its one-byte Identifier has already been assigned to another request. The new RadSec draft keeps the response rejected but refuses to treat every such mismatch as proof that the protected connection has failed.
IETF
The BGP Rule That Went Live Before It Was Finished
The latest GROW security inventory makes a configuration interval visible: on some routers, a permit command can start exporting routes before the rest of the intended policy exists. The change record must cover that interval, not just the finished text.
IETF
The Certificate Expires in 9999. Its Document Binding Can Still Go Unchecked.
A one-signature certificate is designed to accompany one document, one private key and one signing operation. Its proposed X.509 extension narrows scope sharply—but ordinary signature success does not prove that a verifier compared the document with that binding, or that the key…
History
The DSCP Chose a Behavior. The Datapath Was Still a Graph: RFC 3290
In 2002, RFC 3290 gave network managers a way to describe what a Diffserv router could do to traffic without pretending that every vendor built the same box. Its central entity was not a codepoint but a configurable graph of functions.
IETF
The Composite Signature Passed. The Envelope Named a Different Algorithm.
Composite ML-DSA combines post-quantum and traditional signatures behind one interface. Its CMS profile shows why that single verdict still depends on several algorithm identifiers, exact encoded bytes and a content digest agreeing across the envelope.
IETF
OSPFv3 Cannot Retire AH One Router at a Time
A new proposal would remove an old authentication choice from OSPFv3. Its operational test is less about selecting a replacement than about getting every router on a link to accept the same protected packets throughout a change.
History
Rerouting Wasn’t Session Recovery: RFC 3234
When a router failed, the familiar remedy was to find another path. In February 2002, RFC 3234 asked what happened when the failed device did more than route. A middlebox could hold state that the session needed, so restoring reachability did not restore the conversation.
History
The End-of-Frame Bit Could Be the Packet That Vanished: RFC 3189
The receiver wanted one clean fact: the picture was finished. RFC 3189 put that news in the marker bit of the last packet, then warned that the packet carrying the news might be the one the network lost. The convenient signal could shorten delay. It could not be the authority for…
IETF
The Final Filter Was Correct. The Update Still Leaked a Route.
Revision 03 of an IETF GROW draft turns BGP security from a static configuration question into a transaction question. A router may begin and end with the intended policy while briefly enforcing something else between commands.
IETF
The Issuer Signed the Envelope. It Still Could Not Swap the Agent Key.
Revision 02 of the AIC JSON Web Token profile places a principal-signed delegation inside an issuer-signed credential. The nested design matters because an issuer may certify the carrier without acquiring the right to substitute the agent key chosen by the principal.
IETF
The Agent Signed the Challenge. Its Embedded Key Still Wasn't Authority.
OpenA2A AIP revision 03 separates a mathematically valid response from an authenticated agent. The decisive control is not whether the supplied key verifies the signature, but who is entitled to bind that key to the claimed identity.
IETF
Every Link Verified. The Timeline Could Still Be Shorter.
AER-1 revision 09 gives an AI-agent job a verifiable internal history. Its more important contribution is admitting why that history still needs a witness outside itself.
IETF
The Chain Proved a Clock Was Wrong. It Did Not Name the Culprit.
Roughtime can seal an inconsistency into portable evidence without pretending that cryptography has already decided whom to distrust.
IETF
No One Reported Progress. Why Did the Task Look Complete?
A task with entities but no progress reports should preserve the absence of evidence, not turn an empty column into a green completion state.
IETF
The Receiver Saw the Failure. Did the Multicast Head Hear It?
In a multicast network, the receiver that notices silence is not necessarily the node that can act on it. The latest BIER BFD draft makes the route back to the sender—and the proof that an alert was matched to the right session—a separate operational question.