turadg

Sounds like Magic Wormhole (https://magic-wormhole.io/). The readme acknowledges Warner at the end.

Key differences seem to be resumable transfers and proxy support.

Another is a single binary, whereas MW is in Python, but there’s now https://github.com/magic-wormhole/magic-wormhole.rs

I appreciate that the project acknowledges its inspiration. I wish it were more common for projects to provide detailed comparisons with similar ones for people trying to choose.

Funny that “GitHub should really support stacked diffs” led the Graphite team to a space colonization co.

2020: Leave Meta and start a company.

2020–2021: Spend ~18 months looking for PMF.

2021–2025: Build Graphite around stacked PRs, code review, and merge queues.

2025: Get acquired by Cursor because AI makes code review the bottleneck.

2026: Cursor gets acquired by SpaceX because Elon.

Not a startup arc I would have predicted from `gt stack submit`.

Broken promise 1 (Automatically generating CHANGELOGs) is spot on, especially in a monorepo with multiple released artifacts. A single commit could be a breaking change in one package and refactor in another. But the changelog tools see one commit and include it in both changelogs.

Another issue is that once the commit lands on trunk, you can’t revise the entry without editing history. You have to remember to fix it after the changelog is generated.

Changesets (https://github.com/changesets/changesets) is a much better approach. We adopted it in Endo (https://github.com/endojs/endo/tree/master/.changeset) and it’s been a clear improvement.

The new *deno pack* command is a nice addition for safe and simple packaging.

For those using Node.js, a similar single command is available with https://www.npmjs.com/package/ts-node-pack

Now that Node.js supports importing .ts modules, more repos can use them without a build step or putting any build artifacts in the checkout.

Git worktrees are awesome but they broke my workflow in a couple ways:

Resuming work. I used to `j <reponame>` then `gco <branchname>`. Now if I do that I get an error about the branch being checked out already in another worktree. I realized the branch names are pretty unique across repos so I made ` jbr <branchname>` that works from anywhere.

Jumping within repo. The other kink was when I wanted to focus on a particular package I’d do `j <subdir>` and it would usually be unique enough to jump to the one in my current checkout. But now I have dozens of concurrent checkouts and have to pick, even though I’m already in the repo. So `jd <subdir>` does like autojump or zoxide but only within the current checkout.

To power those shell functions I made a “where” extension for Git.

https://github.com/turadg/git-where

It’s working out nicely!

an allegory of SaaS in the AI age

Execa package works nicely for that. Zx has a good DX but is YA runtime.

https://github.com/sindresorhus/execa/blob/main/docs/bash.md

China begins assembling 2,800-satellite network of AI supercomputers

I was on a date when she asked “do you like magnetic fields?”. “Yes!” and we were both smitten.

Took us a few conversational turns to realize we had different capitalization. (This was in the old days when courting was mostly by speaking in person.)

I enjoyed the name Euphoria State University for his fictional model of UC Berkeley.

To combine this news with the AI zeitgeist, consider reading his 2001 novel, Thinks….

« Gen Zers, it turns out, are "not on a linear journey to evaluate the veracity of anything." Instead, they're engaged in what the researchers call "information sensibility" — a "socially informed" practice that relies on "folk heuristics of credibility." In other words, Gen Zers know the difference between rock-solid news and AI-generated memes. They just don't care. »

https://archive.is/BZIU8#selection-1855.0-1867.1