miniblog.

I'm increasingly treating terminal sessions as short lived: starting a fresh terminal tab keeps scrollback short and easy to search. I'd love to have a terminal emulator that allows me to skip over output of specific verbose commands when searching.

What types should you infer when code is invalid? In Garden, + only applies to integers. What should the type of "x" + "y" be? (1) Int: This produces cascading errors. (2) String: We've already emitted an error anyway. (3) Bottom. Opinions?

I'm impressed and slightly surprised that I can browse the web even when my local library's wifi blocks detectportal.firefox.com ("Download sites are banned"). It seems that Firefox only requires the website to be redirected when there's a captive portal.

I've never seen a good design philosophy on how to sort compiler errors for the best UX. Do you order by line, or by severity? Do you put the most actionable at the end (because that's most visible when the terminal scrolls) or at the beginning?

There are labels that make technology ideas sound more exciting. For example, 'augmented intelligence' was used early in the history of computing. (A pocket calculator would qualify but it's less cool.) Is there a good term for this phenomenon? It's hype that inspired projects.

Looking at how my servers update themselves and the recent Copy Fail vulnerability, I'm considering an automated weekly restart to make sure services and kernels are up to date. It'd also ensure that services are correctly configured to start on boot. Any thoughts? Feels crude.

It's funny how languages can offer multiple forms of syntax, but formatters standardise to a single form. E.g. single vs double quotes in JS, optional semicolons in JS, different ways of grouping imports in Rust. Should new languages be more syntactically opinionated?

Difftastic 0.69 is out! This is a pretty big release: * Semantically insignificant commas are now ignored * Difftastic now respects .gitattributes * Added assembly support and improved 13 other languages

I sometimes see blog posts lamenting how inconsistent UIs are in modern apps. When did consistency peak? Windows 3? After the release of Aqua for macOS? The era when Bootstrap was the default CSS framework for many web developers?

It's confusing how 'REPL' refers to two different UIs: A CLI like IPython or Julia, where things like multiline editing and syntax highlighting need careful implementation. An editor protocol like SLIME or CIDER where multiline editing is a total non-issue.

One big challenge of open source is that the number of users (and bug reports) is entirely independent of the number of maintainers. Both users and maintainer capacity can fluctuate wildly.

I've been playing with Obsidian and having a great time. It's fundamentally a .md editor but it has so many affordances that it feels different. Link autocompletion, highlighting backreferences, polished mobile app. A lot of teaching resources focus on folder structure, oddly.

AI agents (OpenClaw, Hermes Agent etc) seem to be increasingly standardising their design. You have a UI for the markdown files writable by the agent (sometimes just Obsidian) plus a set of CLI tools for performing actions. I've seen so many new CLIs recently too.

LLMs are surprisingly good at reducing crash samples. I've had success with "this project crashes my static analysis tool with the following command, try to shrink the repro whilst preserving the crash. Commit each smaller version to a branch".

Has the JS ecosystem reached a point where people agree roughly what a good architecture looks like? I don't see as many blog posts complaining about new framework churn and I'm wondering why.