RFC 863 – Discard Protocol (1983)

https://datatracker.ietf.org/doc/html/rfc863

38•gurjeet•3mo ago

Comments

bawolff•3mo ago

The networked webscale database we've all been waiting for.

bux93•3mo ago

HeadlessChild•3mo ago

/dev/null as a service.

ZeroConcerns•3mo ago

Yes, simpler times and such. And I get the feeling someone is about to discover RFC 864, which is even more fun (as in: a DDOS amplification vector of note, but this stuff actually was useful for a while...)

xg15•3mo ago

https://www.rfc-editor.org/rfc/rfc864.html

> UDP Based Character Generator Service

> When a datagram is received, an answering datagram is sent containing a random number (between 0 and 512) of characters (the data in the received datagram is ignored).

> The service only send one datagram in response to each received datagram, so there is no concern about the service sending data faster than the user can process it.

Oof...

Yeah apparently the idea that the "user" might not be the real sender wasn't yet well-known.

Simpler times indeed.

xg15•3mo ago

I get the TCP-based one, as the service would still complete the connection handshake, send ACKs, etc - but the UDP one seems indistinguishable from simply dropping the packets.

Maybe back then the designers still expected that hosts would always reply to unwanted packets with an ICMP error, so silently dropped packets were expected to be rare and always indicators of a connection fault?

Though I guess we can proudly say today that UDP:9 is the most widely deployed service on the internet...

adrian_b•3mo ago

Yes, indeed it was expected to reply with ICMP errors when receiving packets to unused ports and the necessity of firewalls was not predicted, because the "barbarians" were not using the Internet yet.

Nowadays the well configured servers send ICMP errors only for the traceroute port range and the badly configured servers, which are more common, do not send any ICMP errors for unused ports.

dfe•3mo ago

It's common for wake-on-LAN clients to send UDP packets to port 9 to make sure they get discarded. This is particularly useful if using a multicast or broadcast destination, which is often the case because the ARP entry will have been discarded by the time you need to send the packet.

The hardware that looks for the magic packet ignores the framing.

I certainly wouldn't run a TCP discard service, but making sure that UDP packets to port 9 do not result in any ICMP port unreachable response, or any other response, is a good practice.

Show HN: LocalGPT – A local-first AI assistant in Rust with persistent memory

Haskell for all: Beyond agentic coding

SectorC: A C Compiler in 512 bytes (2023)

Speed up responses with fast mode

Software factories and the agentic moment

Brookhaven Lab's RHIC concludes 25-year run with final collisions

Hoot: Scheme on WebAssembly

Stories from 25 Years of Software Development

LLMs as the new high level language

First Proof

Vocal Guide – belt sing without killing yourself

Show HN: I saw this cool navigation reveal, so I made a simple HTML+CSS version

FDA intends to take action against non-FDA-approved GLP-1 drugs

Al Lowe on model trains, funny deaths and working with Disney

Vouch

Show HN: Axiomeer – An open marketplace for AI agents

Show HN: A luma dependent chroma compression algorithm (image compression)

Total Surface Area Required to Fuel the World with Solar (2009)

Start all of your commands with a comma (2009)

The AI boom is causing shortages everywhere else

Microsoft account bugs locked me out of Notepad – Are thin clients ruining PCs?

I write games in C (yes, C) (2016)

Selection rather than prediction

Learning from context is harder than we thought

OpenCiv3: Open-source, cross-platform reimagining of Civilization III

The F Word

Where did all the starships go?

Unseen Footage of Atari Battlezone Arcade Cabinet Production

The silent death of good code

Reinforcement Learning from Human Feedback