Topic
Network-resource Evidence
Within the Topic facet, Network-resource Evidence topic intelligence connects articles that share a specific subject, signal focus, or monitoring theme. The page gives readers a richer path through related reporting, source evidence, market actors, and infrastructure implications, with enough context to understand why the topic matters across company movements, governance decisions, regional exposure, and operational risk. Readers can compare recurring signals, affected organisations, public evidence, market context, service continuity, procurement, competition, compliance, and strategic planning questions behind the subject instead of stopping at a thin list of matching articles. It explains what the topic covers, which infrastructure actors or policies are involved, what evidence supports the coverage, and why the subject may matter for operators, customers, investors, and policy readers.
History
Why Congestion Is Marked per Packet but Read per Byte: RFC 7141
In 2014 the IETF drew a boundary between making a congestion signal and interpreting its weight. A router should not give a small packet a softer chance of being marked; a transport may still count a marked packet in proportion to the bytes it carried.
History
The Packet Carried No Frame Count. The Receiver Had to Divide: RFC 3952
RFC 3952 made an economical promise: a receiver could recover the number of iLBC speech frames without a count field. The bargain saved bits, but it moved authority into the session description and made one printed divisor matter far more than it looked.
History
Four Zero Octets Separated IKE from ESP. They Authenticated Neither: RFC 3948
RFC 3948 let negotiation, protected data and NAT maintenance share one UDP mapping. Its smallest rule did the first sorting: four zero octets meant “send this toward IKE,” while a nonzero first word could be an ESP SPI. The rule selected machinery; it did not supply the security…
IETF
The Pool Returned Another Server. It Did Not Restore the Session.
The first server stopped answering after a request left the client. `GETNEXTSERVER` did exactly what the pool expected: it marked that Pool Element unreachable and returned another address. Yet the decisive fact was still missing. No one knew whether the old server had committed…
IETF
The Registry Assigned a Router Alert Value. The Router Still Chose Whether to Look.
The packet carried a valid Router Alert option and a value recorded by IANA. It reached a transit router whose operator had configured the option to be ignored. The packet continued as ordinary traffic; no control-plane handler saw it, no reservation state changed, and no…
IETF
The Reflector Returned the Evidence. It Kept No Measurement Report.
TWAMP's Session-Reflector receives a test packet, timestamps it, creates a response and sends the relevant packet evidence back. It does not retain packet-level results for a Fetch-Client, because TWAMP has no such role. RFC 5357 therefore places measurement custody at the…
IETF
The KDC Sent a Preferred Curve List. The Message Was Not Authenticated.
The client proposed elliptic-curve parameters. The Key Distribution Center rejected them and returned an ordered list of alternatives. That list could guide the next authentication attempt, but RFC 5349 states that the Kerberos error carrying it was not integrity protected. The…
IETF
The KDC Suggested a Curve. The Error Message Had Not Been Authenticated.
The client offered one elliptic curve and the Key Distribution Center declined it. The reply looked authoritative: error 65, followed by a preference-ordered list of acceptable parameters. The client could select a curve and try again. RFC 5349 places one hard limit on that…
History
The Code Point Was Private. Its Top Bits Still Controlled Every Unknown Router: RFC 3936
RFC 3936 opened RSVP number space to standards, experiments and private extensions without pretending that a number was merely a label. The first bits of an entity class already told an old node whether to reject the message, discard the entity or carry it onward untouched.
IETF
The Command Succeeded. The Fax Method Had Disappeared.
The same MGCP `ModifyConnection` can produce two opposite operationally useful results. Leave the fax option unstated and the gateway may accept the command even though the previously required T.38 procedure will no longer run. Repeat `t38` explicitly and the gateway must reject…
IETF
The Federation Received the Full Presence Document. Nobody Proved It Filtered the Next Copy.
The optimization was elegant: instead of sending a different presence document for every remote watcher, the home network sent one rich document across the federation and entrusted the receiving network to remove what each watcher was not allowed to see. Bandwidth fell. The…
History
The RFC Was Missing. The DHCP Option Was Not.
RFC 3679 returned DHCP option numbers that had no durable use, but it refused to treat an empty RFC shelf as proof that every code was free. PXE and Apple implementations were already using some undocumented numbers. Sixteen years later, a captive-portal experiment would show why…
History
The Rule Had a Sunset Date. Its Evidence Report Was Optional: RFC 3933
RFC 3933 gave the IETF a clock for temporary process changes. The rule had to expire or be reconsidered, yet the problem statement, success criteria and written account of what happened could remain optional. A sunset constrained authority; it did not manufacture evidence.
IETF
The Call Completed. ENUM Had Not Supplied the Route.
The operations dashboard marked the call as successful. A SIP `200 OK` had arrived, the caller heard ringing, and the session moved into the billing system. Yet the route ledger told a different story: the ENUM query had failed, the softswitch had returned to its vendor-specific…
History
The Message Looked Like a Document. The Protocol Had Already Split It Apart: RFC 3930
RFC 3930 warned that a network message can look like one digital document to a person while the protocol treats it as many fields with different histories, destinations and security rules. The mistake was not aesthetic. It could make extensions fail, signatures brittle and…
IETF
The Manager Learned the Engine ID. It Still Had No Permission to Change Anything.
The discovery reply looked like the missing credential. A management application had reached an SNMP endpoint, asked under a well-known local context, and received the identifier of the engine behind it. The value was stable enough to address later scoped requests. It might even…
IETF
The CSV Said the Poll Succeeded. The Trap Could No Longer Be Reconstructed.
One row looked complete: timestamp, source, destination, SNMP version, operation, request ID, zero error status and a list of variable bindings. It was compact enough to scan, join and graph. But when the question changed from “how many responses succeeded?” to “what did this…
History
The Name Crossed Protocols. Its Parameters Did Not Move by Themselves: RFC 3555
RFC 3555 did not merely give RTP codecs names that could be printed in another protocol. It specified a custody chain: each piece of a media-type expression had a different destination in SDP, and the payload specification—not the registry—kept authority over what the pieces…
IETF
The Parameter Had an IANA Row. The Call Still Had No Route.
RFC 5341 gives `tel` URI extensions a controlled namespace and a durable reference. That is vital coordination. It is not a runtime receipt for the URI, the number, the network or the call.
History
The Name Was Permanent Because It Refused to Name the Current Value: RFC 3553
A registry can move without forcing every protocol document and implementation to rename what it means. RFC 3553 achieved that continuity by giving the name less authority than people are tempted to read into it.