Hacker News new | past | comments | ask | show | jobs | submit | jamesponddotco's comments login

I've been working since I was 12, and working from home since I was 19 (32 now). I can't imagine ever going back to an office, especially now that my wife is pregnant. I'd rather learn a new profession than doing that.


You're lucky you're in a position where you have that choice. Many will be forced to find a different company that offers remote or will actually suck it up back into the office.


Wrote this for work the other day and thought people might be interested. The tool can review and draft technical documentation, as well as generate alt tags for images.

It supports both OpenAI and Anthropic's models right now, but PRs adding support for other providers are welcome :)


I've to write alt tags daily and I still suck at it, since I suck at describing things (which is weird, since I write documentation every day). I might start attending my friend's D&D sessions just to improve on that.

For now, I wrote a tool[1] that uses AI to do the job for me.

[1]: https://git.sr.ht/~jamesponddotco/allalt


Wrote something similar[1] using Go a while back, but I love that this one only uses shell tools. Extremely useful when working with LLMs for small to medium projects.

For bigger ones on the other hand, I wish I could just get the exact context I need. Maybe something like only the files open in vim?

[1]: https://git.sr.ht/~jamesponddotco/llmctx


> Does anyone know any good FOSS projects that were actually funded by this?

SourceHut[1] comes to mind.

[1]: https://sourcehut.org/


I don't know about others, but I still use it for the same reason I still use XMPP: still works fine and is less annoying than proprietary solutions.

As for encryption, I use OMEMO with XMPP for private conversations, but I use IRC for public forum conversations, so TLS is enough for me. Think Hacker News, but in real-time.


I make playlists on TIDAL and then download the playlist. Or favorite albums and whatnot, and download them.


Another vote for Tumbleweed here.

Heck, the thing is so stable I use it on production servers. The combination of their automated QA plus your own QA on staging is the bomb.


Pretty positive, much better than my experience with OpenAI's models. The killer feature for me is their prompt generator[1], which you can use to create system prompts or improve user prompts. As I said in another thread, the generator is tuned to generate prompts better suitable for Claude, which improves responses.

You can see examples of the prompts it generates in the repository linked below, including a customized version of their prompt generator[2]. This feature significantly improved my experience with LLMs in general, both Claude and local ones, and I now use Claude 3.5 Sonnet for pretty much all my coding, mostly in Go and, lately, Rust.

I mostly use it to improve existing code or to get started, not to generate entire code bases, so I can't say much about that.

I found it lacking in shell scripting though. Nine out of ten times I need to fix the shell script it generates, to a point where I just gave up trying and went back to writing them from scratch.

Can't wait to see how much better Claude 3.5 Opus is, though.

[1]: https://docs.anthropic.com/en/docs/build-with-claude/prompt-...

[2]: https://sr.ht/~jamesponddotco/llm-prompts/


I left them a long time ago, as I felt—and still do—like the internet was too dependent on them. None of my projects ever needed DDoS protection though, so I mostly used them for DNS and caching at the edge.

If I ever needed DDoS protection it'd be nice to know of alternatives, if anyone has any to recommend. My provider—Hetzner—provides protection to some degree, but I'm sure a big attack would lead to black-hole routing.


Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: