Remote

Josh Triplett

@josh@social.joshtriplett.org

Free and Open Source Software developer. #rustlang developer, working on the language, library, and Cargo. he/him

1039 Followers

81 Following

10 Posts

Homepage:

https://joshtriplett.org/

GitHub:

https://github.com/joshtriplett

GitHub Sponsors:

https://github.com/sponsors/joshtriplett

Posts

Josh Triplett @josh@social.joshtriplett.org · Apr 29, 2026

Replying to @chergert@my.devsuite.app

Oh, I didn't know that! I assumed that setting that in Ptyxis would cause it to always treat Ctrl-C as copy, rather than only if text is selected. I'll give that a try the next time I make an attempt to switch from gnome-terminal to ptyxis.

Josh Triplett @josh@social.joshtriplett.org · Apr 28, 2026

Replying to @chergert@my.devsuite.app

Honestly, I wouldn't use it at all if not for the fundamental conflict between the standard copy-paste keyboard shortcuts and terminal Ctrl-C. I use it because I need to, because it's more usable than Ctrl-Shift-C; I don't use it because I *want* to. If we made it so that terminals interpreted Ctrl-C as copy if text were selected, and only sent Ctrl-C through as an interrupt if text weren't selected, I'd happily switch to that and abandon middle-click paste.

Josh Triplett @josh@social.joshtriplett.org · Apr 25, 2026

Replying to @pid_eins@mastodon.social

I wouldn't say "happily".

Rust standard library folks are very well aware of the ideal of doing fd-based operations whenever possible, and we'd love to. Linux is doing great work on adding ways to do everything one might want to do using fds. However, we can't force people to run on exclusively modern Linux, as opposed to old Linux or other OSes. And it's much more challenging to design portable interfaces around fds without accepting capability limitations or lowest-common-denominator.

We could probably make an extremely capable interface, if we stuck most of it in std::os::linux, and had some of it fail if run on older Linux.

Josh Triplett @josh@social.joshtriplett.org · Apr 16, 2026

Diseases that are contagious before people are symptomatic are especially hazardous and epidemic-prone. The epidemic is time-delayed from the initial outbreak. LLM usage begets more LLM usage, and LLM-written technical debt takes time to become symptomatic. Exponentials are harder to see from the middle.

Josh Triplett @josh@social.joshtriplett.org · Apr 02, 2026

Replying to @Captcha@theforkiverse.com

I'd want to be a bit broader than "must be reverse-chronological" (e.g. let's not require a legislative update to add creative things like "show me a subset list of people I've curated" or "show me posts with this tag I specifically bookmarked"), and I don't especially think Section 230 is the right tool for this, but the general *concept* of "restrict algorithmic-optimization-for-engagement-and-enragement" seems sound.

Josh Triplett @josh@social.joshtriplett.org · Apr 01, 2026

Replying to @Casey@theforkiverse.com

I think calling it a "teen mental health crisis" is giving teenagers too little credit and adults too much. Many of the same features that addict and abuse teenagers do the same to adults. I don't think we should view this through a lens of "child safety" (with all the political connotations that has), when it affects everyone. I think we're currently in the period much like the pre-"cigarettes are known to be addictive" one. *Post* "there's a warning on every pack", it's reasonable for people to take responsibility for their own usage (as long as it isn't around others). *Pre* "there's a warning on every pack", it can't be written off as "personal responsibility".

Josh Triplett @josh@social.joshtriplett.org · Feb 13, 2026

In response to many claims that LLMs help with polishing writing: Homogenization is for milk, not for writing.

Josh Triplett @josh@social.joshtriplett.org · Nov 11, 2025

Replying to @esoterra@hachyderm.io

Sure, it can, but we should standardize the interface for an RNG.

Josh Triplett @josh@social.joshtriplett.org · Nov 11, 2025

Replying to @sunfish@hachyderm.io

Counterexample: what if you want to run a game with a replayable RNG so that you get deterministic replays? I think we need two separate RNGs: "replayable" (can be seeded) and "secure" (cannot be seeded). Crypto should use the latter, game RNGs and similar should use the former.

Josh Triplett @josh@social.joshtriplett.org · Aug 03, 2025

Replying to @next@fedi.copyleft.org

I'm interested in contributing to copyleft-next. I've got extensive experience with licensing and FOSS and copyleft, going back decades, including the nuances of GPLv2/GPLv3 and the GPLv3 drafting process. I'd love to help. Could I please get an account (without having to use OID/SSO)?

Josh Triplett

Posts

Remote instance