August 1, 2026
Git good or Git gone?
OpenAI's work on Git for large repositories
OpenAI tries to make giant Git projects faster, and readers say the write-up is a maze
TLDR: OpenAI is working on making Git, the software developers use to manage code, much faster and safer for giant projects. But the biggest reaction from readers was a harsh one: instead of celebrating the upgrade, they roasted the post itself for being painfully hard to understand.
OpenAI dropped a deep-dive on speeding up Git, the tool developers use to track changes in code, especially for huge projects where even simple actions can feel painfully slow. The actual improvements sound useful: faster saves, fewer unnecessary rewrites, safer downloads, and big cuts to waiting time when handling massive code collections. In plain English, they’re trying to make one of coding’s most important tools less of a slog when projects get enormous.
But the real fireworks came from the community reaction, where the loudest takeaway was not “wow, impressive,” but basically: what am I even reading? The top comment, blunt and brutal, called the post “unreadable,” and that mood hangs over everything. Instead of cheering the engineering win, readers seemed stuck on the fact that the update reads like a wall of mysterious commands and internal notes. It’s classic internet drama: one side sees serious behind-the-scenes work that could help developers everywhere, while the other sees a post so dense it might as well be encrypted.
That clash gave the whole thing a weirdly funny energy. The subtext in the thread is almost a meme in itself: OpenAI may be making coding tools faster, but apparently not the explanation. It’s the oldest tech-community joke around — brilliant fix, terrible vibes. For non-experts, the important part is simple: the code-management plumbing is getting faster and safer, yet the public reaction turned into a roast of how impossibly hard this all was to read.
Key Points
- •OpenAI’s update groups Git large-repository work into patches already merged upstream, patches in upstream review, and work still in progress.
- •A merged `git stash` change avoids unnecessary sparse-index expansion for in-cone paths and reports a performance improvement from 18.87 seconds to 0.06 seconds.
- •A merged `pack-objects` change enables reachability bitmaps and delta islands during path-walk, with fallback to ordinary traversal when needed.
- •Under-review patches address safe concurrent packfile-URI downloads, duplicate pack object recovery, `no-ref-delta` negotiation, combined geometric repacking with cruft packs, and correct incremental multi-pack-index layering.
- •In-progress work targets faster `git status`, avoiding unchanged index rewrites in checkout/restore, fixing a Coccinelle-related static analysis slowdown, and tracing external packfile-URI downloads in fetch-pack.