Content Type
Research
Within the Content Type facet, Research intelligence gathers BTW.MEDIA articles that share the same editorial format, helping readers compare briefings, profiles, risk notes, market analysis, and event coverage without mixing different kinds of evidence. The page explains how this content type frames internet infrastructure events, company movements, governance decisions, operational signals, and public evidence across the site. Readers can compare which actors or infrastructure systems appear most often, how source quality changes interpretation, and whether the material is a durable profile, a time-sensitive event, a strategic market signal, or a governance development. The result is a useful search page for operators, investors, customers, analysts, and policy stakeholders who need to understand the consequence, timing, and evidence behind similar article formats.
Global Cloud Services Trends
SailPoint’s AI ARR needs a renewal-status column
SailPoint has put a number on its new AI business: more than US$70 million of annual recurring revenue. The number is useful precisely because it is not yet a revenue line. Its definition can keep a contract in the count after expiry while a renewal or replacement deal is still…
History
Anja Feldmann and the Traffic Matrix No Single Link Could Reveal
A red link on a network map looks like an answer: this is where the load is. For Anja Feldmann and her collaborators, it was only the start of the question. The counter could not say which demands had converged there, which routing decisions had brought them together, or where…
IETF
Max-Forwards Counts HTTP Hops, Not Organizational Authority
Max-Forwards gives an HTTP client a small diagnostic budget: a TRACE or OPTIONS request may cross only so many forwarding steps before an intermediary must answer. That counter is useful precisely because it is narrow. It does not reveal how many companies are involved, certify…
History
Jerry Saltzer and the Link That Carried a Name but No Permission
A path can tell a computer where a file is without answering whether the person following it may read the file. Jerry Saltzer’s account of protection in Multics turned that separation into an operating rule—and exposed how much revocation, audit and institutional power can be…
IETF
Accept-Patch Advertises Patch Formats, Not Permission to Modify
A server can tell clients which patch-document languages it understands without deciding who may change a resource. RFC 5789 gives that discovery statement a name, Accept-Patch, and keeps capability, format semantics, current state and write authority as separate questions.
CASE FILE
The Stream Was Created. The Broadcast Was Not: RFC 9725
A WHIP endpoint can return `201 Created` while the programme remains black. RFC 9725 makes low-latency contribution easier to start; it also makes it essential to distinguish the HTTP resource from the media path, the production chain and the audience result.
History
Eve Schooler and the Invitation That Did Not Carry the Conversation
The most consequential line in an Internet call is not the one carrying a voice. It is the one that finds the other person, proposes a session and then gets out of the way. Eve Schooler’s path from experimental conference control to the early Session Initiation Protocol helps…
IETF
Content-Location Describes a Representation, Not Where the Client Must Go
An HTTP response can identify the resource corresponding to the document it carries without changing the address that was requested. RFC 9110 calls that field `Content-Location` and makes the boundary explicit: it is representation metadata, not a replacement target or an…
CASE FILE
RFC 9727 Put the APIs in a Catalog. It Did Not Put the Catalog in Charge
The most dangerous API inventory can be perfectly formatted. Its links validate, its cache is warm, and its obsolete interface is gone from the list. None of those facts says what is still listening on the network.
IETF
103 Early Hints Can Start a Fetch, Not Settle the Response
A web server can reveal part of its likely answer before it has decided the answer itself. RFC 8297 makes that useful without making it authoritative: a 103 response can fund preparation, while the final response alone settles what the request produced.
History
Lixia Zhang and the Data Packet That Did Not Need Its Original Server
The revealing moment in Named Data Networking is not when a packet reaches a machine. It is when an Interest is answered by a copy found somewhere else—and the receiver can still ask who signed the data, whether that key was entitled to sign this name, and whether a fresher…
IETF
A Problem Type URI Is an Identifier, Not a Remote Command
An API error can name a stable kind of failure without giving the name control over the client. RFC 9457 draws that boundary precisely: the problem type URI identifies semantics; the response, the client’s policy and separately defined evidence determine what may happen next.
IETF
UUIDv7 Is Time-Ordered, Not a Causal Receipt
UUIDv7 places time at the front of an identifier so new values tend to sort near one another. That is valuable for indexes and rough chronology. It does not mean the identifier can testify that event A caused event B, that two machines agreed on time, or that the holder has any…
History
Abhay Bhushan and the Uniform Interface That Stopped at the Pathname
Before a file could cross the early network, its sender and receiver first had to admit how unlike they were. Abhay Bhushan’s work on the first File Transfer Protocol turned that awkward fact into a design method: discover capabilities, standardise a workable layer of operations…
IETF
Cache-Status Is a Chain of Claims, Not a Cache Verdict
A single HTTP response can carry several Cache-Status members, each written by a different cache and each describing only its own handling of the request. Read that ordered list as a provenance record and it becomes useful. Compress it into one global “hit” or “miss” and the…
History
Cynthia Dwork and the Cost That Could Not Be Reused
Before proof of work became a cryptocurrency term, Cynthia Dwork and Moni Naor treated computation as an admission price for email: small for one legitimate message, punishing when multiplied across a bulk campaign, and useful only if the same payment could not be spent twice.
IETF
RFC 9730 and the Evidence Boundary Between Distributed Recovery and Centralized Control
RFC 9730 matters less because it settles an architectural contest than because it makes a mixed-control reality explicit: transport networks can combine distributed GMPLS behavior with centralized coordination, and the operational challenge is therefore to distinguish who…
IETF
HTTP must-understand Is a Two-Directive Upgrade, Not a Magic Word
Put `must-understand` and `no-store` in the same HTTP response and two generations of cache can make different, deliberate choices. The older one falls back to not storing; the newer one may take a narrower path only after it proves that it implements the caching rules of the…
History
Raj Jain and the One Bit That Needed Two Time Filters
DECbit is remembered as a one-bit congestion signal. The more consequential idea was the timing around it: routers and endpoints each had to decide what a stream of momentary observations meant before anyone changed the traffic load.
IETF
Requesting an HTTP Digest Does Not Create an Integrity Contract
An HTTP client can state exactly which digest algorithms it would prefer and still receive a response with another algorithm—or no digest at all. RFC9530 makes that latitude deliberate. Operational integrity begins only after the recipient inspects what actually arrived…