#Latency

2025-06-15

#LibreQoS Launches #Bufferbloat Testing Platform:

libreqos.io/2025/06/13/bloat-t

Traditional speed tests only tell part of the story. We tried to address @mtaht's points from “What’s Wrong with Speed Tests” article to make a better #speedtest that measures #WorkingLatency, #LatencyUnderLoad, and #jitter.

#DaveTaht #speedtests #latency #RFC8290 #ISP #WiFi #broadband #OpenSource #FLOSS #FQ_CoDel #sch_CAKE #WISP #FISP #QoE #QoS #QualityOfExperience #QualityOfService #InternetServiceProvider

N-gated Hacker Newsngate
2025-06-08

Oh look, scientists claim they can sneak into T cells to tackle latency—because if there's anything more efficient than human cells, it's that suggests you upgrade your browser. 🙄🖥️ Maybe next they'll solve world hunger with a new app that only works on Netscape. 🍽️✨
nature.com/articles/s41467-025

2025-06-05

This is Major FFMPEG has merged WHIP Muxer for sub second latency streaming!!!

WHIP is the acronym for WebRTC-HTTP Ingestion Protocol. It is a low-latency live-streaming specification building off WebRTC.

Read more about it here

phoronix.com/news/FFmpeg-Lands

#OpenSource #FFMPEG #WHIP #WebRTC #FFPlay #streaming #latency #low #programming #POSIX #Linux

The image shows a screenshot of a news article on a mobile device. The article is titled "WHIP Muxer Merged To FFmpeg For Sub-Second Latency Streaming," written by Michael Larabel in Multimedia on June 4, 2025, at 08:00 AM EDT, with 6 comments. The article discusses the merging of a big project into FFmpeg to provide a WHIP muxer for sub-second latency streaming. It explains that WHIP stands for the WebRTC-HTTP Ingestion Protocol, a low-latency live-streaming specification building off WebRTC, which uses HTTP for exchanging initial information and capabilities and STUN binding to establish a UDP session. Encryption is supported, and due to WebRTC, mandatory with WHIP, and audio/video frames are split into RTP packets. The WebRTC-HTTP Ingestion Protocol is an IETF standard for ushering low-latency communication over WebRTC to help with streaming/broadcasting uses. The screenshot also includes a code commit interface for FFmpeg, showing "Commit 167e343" with 7 people committing 7 hours ago. The device's status bar shows the time as 22:08, the date as the 31st, and a battery level of 91%.

 Ovis2-8B

🌱 Energy used: 0.320 Wh
N-gated Hacker Newsngate
2025-06-01

💡🤔 Ah, the age-old tale of the that’s somehow both a demon and a sloth. Turns out, running locally is like asking a turtle to win the Indy 500—slow, steady, and hilariously unsuited. 🚗🐢 So, remember kids, if you want your to hustle, scale it or face the wrath of latency! 👴📉
seangoedecke.com/inference-bat

Jean-Baptiste "JBQ" Quérujbqueru@mastodon.gamedev.place
2025-05-30

As a noob about writing games, I'm facing questions about the fundamentals:

Does it make more sense to run my game logic against an absolute timer, and have the renderer run asynchronously, downstream from the game logic, or does it instead make more sense to run the game logic in sync with the renderer?

Running asynchronously is actually much simpler algorithmically, because separation of concerns, but adds a tiny bit of #latency (~8ms average if the logic runs at 60Hz).

2025-05-28

@nielso @_slotek_ Sorry to hear that. #Latency is surely an issue with low frequencies. There will always be a perceptible delay which becomes uncomfortable above 30ms but won't go below that either because of algorithmic latency + ADC latency. Is there less latency on higher strings?
Since it all comes down to algorithms, someday a new one might arrive to solve all our big-buffer problems in the low range ;) That's why I'd go #Cycfi rather than #Roland because of updateability of the actual part that tracks the pitch.

Hacker Newsh4ckernews
2025-05-28
2025-05-21

Running your #homelab - what is more important to you: #latency or #upload

Would you prefer running your lab connected by:

Client Info

Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst