flashtea/
paper

No. 126 · Tuesday, 20 October 2026 · 3 stories

Dev Tools & Releases

GitHub said it may temporarily require authentication for HTTPS clones and fetches of public repositories during periods of unusual traffic

The protection applies to unauthenticated Git over HTTPS on all plans and is meant to protect github.com's availability; affected operations fail with a message saying authentication is required. Authenticated HTTPS, SSH, pushes and the web and API are unaffected, and GitHub says this is not a general authentication requirement for public repositories.

why it mattersAny CI job, Dockerfile or script that clones a public repository anonymously can now fail intermittently, so switch it to GITHUB_TOKEN in Actions or a GitHub App token elsewhere, and treat unauthenticated clones as best-effort.

GitHub Changelog

World0 stories

Ahead

AI for Software Development0 stories

Ahead

  • 28 OctGitHub Universe 2026, San Francisco (28–29 Oct)
  • 29 OctLast day of the grandfathered 20x-Plus allowance for existing ChatGPT Pro $200 subscribers in Codex; drops to 10x from 30 Oct
  • 10 NovAI Engineer Code Summit, San Francisco (workshops 10 Nov, talks 11–12 Nov)
  • 16 NovGitHub Copilot retires Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra

AI Research & Models2 stories

Inception released Mercury 3, a diffusion language model with a 1-million-token context that it says generates 1,068 tokens per second on Nvidia GPUs

The company says the model has about 40 times the parameters of Mercury 1, matches or beats GPT-6 Luna and Claude Haiku 5.5 on its benchmark table while running four to eight times faster, and is available through its API, OpenRouter, Baseten and Vercel at $0.20 per million input tokens and $0.70 output, half price at launch. All numbers are Inception's own; no independent evaluation has been published, and the model trails Haiku 5.5 on coding rows.

why it mattersDiffusion decoding at four-figure tokens per second is now a production option for the latency-bound steps in an agent (classification, routing, voice), so it is worth a benchmark on your own traffic before taking the quality claims on trust.

Inception

MiniMax released M3.1 Flash at $0.15 per million input tokens and $0.60 output, scoring 73.7% on DeepSWE v1.1 with weights promised later

The company reports 61.7 on OSWorld V2, 82% on CyberGym and median output speeds above 170 tokens per second, and says agentic tasks cost about 70% less than on its predecessor. The weights are not yet public and no license has been named.

why it mattersA sub-dollar model within a few points of DeepSeek's V4.1-Flash on agentic coding benchmarks is another candidate for the cheap-subagent tier, with the usual caveat that the scores are the vendor's.

MiniMax

Ahead

Dev Tools & Releases1 story, lead above

Ahead

  • 28 OctNode.js 26 enters LTS; Node.js 27 alpha opens under the new one-release-a-year schedule
  • 04 NovGitHub: first ssh-rsa SHA-1 brownout (second on 9 Dec, removal 13 Jan)
  • 17 NovChrome 158 stable removes XSLT for all users outside the origin trial and enterprise policy; PostgreSQL 19 GA two days later
  • 19 NovGitHub Actions finishes moving the ubuntu-latest label from Ubuntu 24.04 to 26.04 (rollout began 19 Oct; pin ubuntu-24.04 to stay)

Money0 stories

Ahead

  • 21 OctUK September CPI (ONS)
  • 23 OctMoody's scheduled review of France's Aa3 rating (negative outlook); flash PMIs and Japan CPI the same day
  • 28 OctFOMC rate decision; UK Autumn Budget with OBR forecast the same day
  • 29 OctECB rate decision and US Q3 GDP advance estimate; Bank of Japan decides the next day
Bitcoin · weekly close vs 200-week moving average · log scale · last 8 years Price200-week average
$5,000$10,000$20,000$50,000$100,000 20192020202120222023202420252026 $80,809 $67,358
$5,000$20,000 2020202220242026 $80,809 $67,358

Price is 1.20× the 200-week average. Weekly closes, USD, as of 28 October 2026 · data blockchain.com.