861–870 of 913 posts
Claude/gemini helped with design and content meta. NLM for content with official sources.
I am doing a similar with linux internals (to be published soon). ofcourse from the official docs and asking some questions around the concepts i like and forget mostly :-)
Earlier quoted context omitted.
This is really cool! Seems like the key to the prompt it seems is the knowledge of how these tones were created in the first place (your system prompts for the 80s) which requires some actual tone knowledge and leaves all the fiddly bits to the LLM. As someone relatively new to guitar this is intimidating, but also a fun rabbit hole. Additional “expert” prompts for different archetypes and genres could be fun.
thanks.Frankly — the system prompt is the sauce. The LLM has encyclopedic knowledge of every amp, cab, and pedal ever made which still blows my mind, but without guardrails it'll confidently tell you to boost mids on a JCM800 (which is how you disappear in a loud mix i've noticed). That era cheat sheet is all the working knowledge over the years of spectacularly failing to get the right tones. Once someone does that…
(Night shot detector should be faster, instead Python, I would rewrite it in some native language, like Rust, but it will run once per project, so it's not vital.)
It has several functions which a normal RC can't have, e.g. search channel by name, or hide "ABCD" channel, if it has "ABCD HD" pair, to make channel list shorter.
I restore old non-functioning radios with nice design and good audio. I replace the old tech except the speakers with a raspberry add a mic array and package them with an easy to use ubuntu. whisper and the small gemma models made everything so much easier and private. I basically rewrite the whole backend with claude and created a nice first setup experience akin to other smartspeakers. It's amazing how good the sou…
What do you do with the old guts? I suspect plenty of people would be interested in those (me included)
Earlier quoted context omitted.
how Mistral OCR works for you? Is it really better than Tesseract?
IME way better. It may not be the best out there, but it's cheap (2c per page), fast, easy to integrate API, and sufficient for my needs. It does things like describe what's drawn in pictures and shown in graphs, which all helps when searching.
https://github.com/yo-chanan/gmail-imapfilter
I had looked everywhere for a simple system service that supports Gmail, or a chrome extension that supports push, and was able to find neither. Then when I started using Codex, I thought "Here, go automate the setup of the imap-filter service for Gmail."
Earlier quoted context omitted.
Here's what gemini says about running hooks in parallel: > Native Git traditionally executes hooks sequentially. However, you can achieve parallel execution by leveraging dedicated third-party hook managers or using built-in shell background processing .. and that's what git-multi-hook is: a third-party hook manager, that uses shell background processing :)
Just out of curiousity, what are the 8 hooks you run?
- end-with-blank-line: normalize file endings
- find-do-not-commit: do not commit files that include "DO NOT COMMIT"
- lint-code: run `$REPO/scripts/lint` if it exists
- lint-nodejs: run `$REPO/{pnpm/yarn/npm}` lint if `package.json` exists
- lint-shellcheck: shellcheck all the files that have no extensions or end in `.sh` that have a shell-shebang
- lint-swift: run swiftlint
- prevent-commit-secrets: run ripsecrets to avoid committing secrets and credentials
- validate: run `$REPO/scripts/validate` if it exists
By default all of the scripts check the files in the git staging area, but they can also be run standalone to check everything. You can find them here: https://github.com/webcoyote/git-multi-hook. Glad to take suggestions for more.