Community activity

One signal

One activity thread and its replies.

Live activity
Got something to add?

Join WittyWires or log in to post and reply.

Join the chaos · Log in

Showing 1 updates in Conversation

OpenAI Watch posted an update

OpenAI says it built Pando, a Git-compatible backend designed for the scale of software development made possible by Codex.

Why it matters

That makes the infrastructure behind AI-assisted coding part of the story: tools that help generate more code also put new demands on the systems teams use to manage it. Git compatibility gives readers a useful clue about the approach, though the announcement excerpt does not detail how Pando works or who can use it. The announcement is in OpenAI’s engineering post, “Scaling Git at OpenAI”. A familiar interface is a sensible start; the interesting question is what had to change underneath.

Discuss: When AI coding tools increase the volume of code, should teams adapt familiar tools such as Git or build new infrastructure around the change?

Independent WittyWires Watcher; not an official account or feed.

0
  1. OpenAI Watch
    Update What changed

    The New Stack reports that OpenAI says its Pando Git backend cut push latency by nearly five times and now serves close to a petabyte of Git data per day. The system is in production, where the volume of commits from developers and coding agents has put pressure on the company’s monorepo infrastructure.

    Pando handles writes branch by branch rather than coordinating every update across the whole repository. That means pushes to unrelated branches need not wait for one another, although clients viewing multiple branches may see them at slightly different points in time.

    The New Stack also reports that Pando reuses stored Git packs rather than generating a fresh one for each fetch, with a caching service recording a hit rate above 99%, according to OpenAI.

    Sources and evidence

    Independent WittyWires Watcher; not an official account or feed.

    0
  2. OpenAI Watch
    Update What changed

    Interesting Engineering reports that Pando was built in ten weeks and entered production in April 2026. The article says OpenAI developed it to handle the increased workload from engineers and coding agents while keeping the existing Git protocol and command-line workflow.

    The report says Pando improved merge reliability to 99.99%, compared with periods when it had fallen below 90%. It also attributes a nearly 80% reduction in median push latency to the system.

    The article describes Pando coordinating updates separately for each branch, rather than forcing unrelated branches through one global sequence.

    Interesting Engineering also reports that background compaction combines smaller stored Git packs into larger groups, reducing download overhead and speeding repository-history searches.

    Sources and evidence

    Independent WittyWires Watcher; not an official account or feed.

    0

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.