jamesponddotco

I do one question at a time so I don't feel overwhelmed and can answer questions with more details.

I start with an idea between <idea> tags, write as much as I possibly can between these tags, and then go one question at a time answering the questions with as much details as I possibly can.

Sometimes I'll feed the idea to yet another prompt, Computer Science PhD[1], and use the response as the basis for my conversation with the socratic coder, as the new basis might fill in gaps that I forgot to include initially.

[1]: https://github.com/jamesponddotco/llm-prompts/blob/trunk/dat...

[2]: Something like "Based on my idea, can you provide your thoughts on how the service should be build, please? Technologies to use, database schema, user roles, permissions, architectural ideas, and so on."

They are subagents and slash commands, depending on the project. Eventually, I need to come up with a “dotclaude” repository with these and a few others I use in.

Edit: Sorry, I had a brain fart for a second, thought you were talking about other prompts. I prefer to keep those as chats with the API, not Claude Code, but yeah, they might work as slash commands too.

I have multiple system prompts that I use before getting to the actual specification.

1. I use the Socratic Coder[1] system prompt to have a back and forth conversation about the idea, which helps me hone the idea and improve it. This conversation forces me to think about several aspects of the idea and how to implement it.

2. I use the Brainstorm Specification[2] user prompt to turn that conversation into a specification.

3. I use the Brainstorm Critique[3] user prompt to critique that specification and find flaws in it which I might have missed.

4. I use a modified version of the Brainstorm Specification user prompt to refine the specification based on the critique and have a final version of the document, which I can either use on my own or feed to something like Claude Code for context.

Doing those things improved the quality of the code and work spit out by the LLMs I use by a significant amount, but more importantly, it helped me write much better code on my own because I know have something to guide me, while before I used to go blind.

As a bonus, it also helped me decide if an idea was worth it or not; there are times I'm talking with the LLM and it asks me questions I don't feel like answering, which tells me I'm probably not into that idea as much as I initially thought, it was just my ADHD hyper focusing on something.

[1]: https://github.com/jamesponddotco/llm-prompts/blob/trunk/dat...

[2]: https://github.com/jamesponddotco/llm-prompts/blob/trunk/dat...

[3]: https://github.com/jamesponddotco/llm-prompts/blob/trunk/dat...

I open source pretty much everything I work on that is close to finished or finished. Never regretted doing it, but never got anything out of doing it either, aside from the feeling of paying forward.

I guess it really depends on how popular your project gets. I have no idea if my stuff is used or not[1], so regretting is maybe kinda hard?

I’ll keep doing it, though. Might regret it at some point, but I get so much value out of open source, it feels wrong not to.

[1]: Judging by the lack of patches I’d guess my work isn’t used, though.

Inspired by this article by Laurie Hérault[1] which explains how understanding the science of video games helped them "cure" their procrastination, I built a simple CLI tool in Go to print those tasks and added a fun little twist, a "Pokemon roulette", of sorts.

Code is messy since I built it thinking I'd just throw it into a dotfiles repository, but it works, which is what matters. I'll work on improving it after I test the technique and ensure it works for me.

I'm going back to work tomorrow after three months of paternity leave, and will test further to see if the technique really works. Here's hoping.

[1]: https://www.laurieherault.com/articles/a-thermal-receipt-pri...

My brother works with COBOL for a bank here in Brazil, he is young (in his 20s), started before finishing his degree. Pay is poor, hours are insane, he is overworked as hell, and anything “modern”, like git, is out of the question.

He’s trying to learn Go now and modernize himself to see if he can get out. I’m trying to help as much as I can. Hopefully, he’ll land a job somewhere else this year.

Cloudcraft[1][2] can do that with your cloud provider, AWS or Azure. As a bonus, the diagrams also look quite cool.

[1]: https://www.cloudcraft.co/

[2]: I’m part of the Cloudcraft team at DataDog, so obviously, I’m biased.

My mother-in-law did the same, then I cut her off, no more pictures of the kid. She got into the program after that.

I live in the terminal, so I wrote my own prompt ages ago when I started learning Go[1][2], and before that I had a simple prompt in bash.

I like to keep things very simple and fast, so the directory and the git branch is all I need. I wonder if people really use all that information or if they set it up thinking they need it, but then never do.

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

[2]: I should probably update that now that I know a “bit” more Go.

That is correct, there's no alternative front-end that still works. Self-hostable open source ones, that is—you can still find random ones on search engines that aren't open source.

Seems related, so I’ll share here. I wrote an “awesome” list of privacy-focused front-ends[1] for a variety of services. Haven’t been updated in a while, but I figured it’s still valid.

[1]: https://sr.ht/~jamesponddotco/awesome-privacy-front-ends/

And tracking.

I’ll see if I can find it here (Brazil). Thanks!

Any recommendation for a good thermal printer that works with macOS? This thread gave me a few ideas that I might hyper fixate on and then forget about in the span of a day or two.

Personally, I prefer DNSCrypt over DoH or DoT (and run my own DNSCrypt servers), but I see no problem with DoH. It's certainly an improvement over unencrypted DNS.

If I ever sell the house, I’m taking everything with me. Nothing I built is permanent to the house, so aside from having to paint over some 3M tape here or there, it should be fine.

I’d never buy a “smart house” though. I have to set up everything myself, that’s part of the fun.

I have the complete opposite experience, as for me and my wife Home Assistant has been excellent. I mostly use Zigbee though, as it can’t phone home, and that may be related.

Heck, I even set up the HomeKit bridge so my wife can use Siri or the Home app to control everything, and she still prefers the dashboard I built on Home Assistant.

I only have two problems, which aren’t even Home Assistant problems:

1. Some door and window sensors I got from Sonoff become unavailable at times, which is a problem since I use them on the security system. Haven’t had the time or money to replace them yet, but I’ll probably go with Aqara next.

2. While my wife loves the dashboard, I like to thinker with it, and am never satisfied. I’m still looking for a way to expose all devices and controls in a visually pleasing way.

Actually, three. My only cloud integration, which controls my LG air conditioner, recently started to give me problems and I can’t turn the units off from Home Assistant anymore. But I’m sure this will be fixed soon, so, meh.

All in all, I’m very pleased with Home Assistant. Wish they had a better design, but themes fix that, so, overall, very pleased.

If anyone is looking for something similar, but better priced, the Ulanzi TC001[1] with a custom firmware[2] is pretty nice. With the custom firmware, you can integrate it with Home Assistant and do all sorts of things with it.

For example, one idea my wife had was to have one on top of the washing machine displaying the last cycle of the machine based on the energy readings from a Zigbee smart socket, so we don't have to ask each other if the last cycle was a wash or drying cycle.

Will it change your life? No. But for $59 buckets it may add convenience here or there ¯\\\_(ツ)_/¯

[1]: https://www.ulanzi.com/products/ulanzi-pixel-smart-clock-288...

[2]: https://blueforcer.github.io/awtrix3/

As a software engineer, probably JavaScript, as I focus on back-end and don’t have many front-end skills aside from a good hand at design.

In general, though, since I believe I’ll lose my job to AI, I’d like to learn and become an electrician.

The focus is cloud architecture diagramming, but you can use the design tools to diagram bare-metal architectures if you want or anything else, really. Heck, we had an internal game of chess going at some point and someone was rebuilding pieces of New York City using the design tools.

I'm clearly biased since I work at Datadog[1], but I use and recommend Cloudcraft[2] for diagramming. As I don't use the cloud I focus more on the manual design tools, but the automatic stuff are pretty neat.

Recently, though, I've been using D2[3] a lot and really liking it. The diagrams don't look as aesthetically pleasing to my eyes, but being able to design everything with a simple language is pretty cool, and helps with automation.

[1]: Datadog owns Cloudcraft, hence the bias. Plus, I work in the Cloudcraft team, haha.

[2]: https://www.cloudcraft.co

[3]: https://d2lang.com

I love diceware passwords, they’re the perfect way to keep things secure while allowing you to type them on a TV app, for example. Nice to use for the master password of your password manager too, since you can remember a very long one.

I wrote a Go package[1] that generates them (and random ones too) a while ago and also compiled a word list with about 23 thousand words[2]. I had it reviewed by a third-party, and, it seems, bias wasn’t introduced in the generation. There is a CLI app[3] using the package, if anyone is interested in something written in Go instead of Python for whatever reason.

[1]: https://git.sr.ht/~jamesponddotco/acopw-go

[2]: https://git.sr.ht/~jamesponddotco/acopw-go/blob/trunk/words/...

[3]: https://git.sr.ht/~jamesponddotco/acopw-cli

Hi, folks!

I just released v1.0.0 of pkgdex, a simple self-hosted service that provides both a searchable index for your Go packages and vanity import path handling.

It offers full-text search, vanity import path support, naive download statistics, and more. You can see a live version of the service running at https://go.cipher.host/.

Been a while since the last time I wrote a server, so there may be issues here and there, but I hope you folks like it. Constructive feedback is much appreciated and PRs are welcome!

Yep, but I choose to focus on the meme since it makes me chuckle.

I don’t have an answer for everything you are looking for, but I wrote bonk[1] to solve a similar issue, as I needed to ensure users weren’t uploading porn[2]. Maybe you can find an use for it too.

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

[2]: Because I host in Germany.

Interesting! I wrote a prompt for something similar[1], but I use Claude Sonnet for it. I wonder how a small model would handle it. Time to test, I guess.

[1]: https://git.sr.ht/~jamesponddotco/llm-prompts/tree/trunk/dat...

Like many people, I built something similar, llmctx[1], but the chunking feature seems really interesting, I have to look into that for llmctx.

One thing I have with llmctx that I think is missing in yek is a “Claude mode”, as it outputs the concatenation in a format more suitable to provide context for the Anthropic LLM.

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

Why would I care about serverless? I love managing and working with bare metal servers.

I usually encrypt with LUKS and then delete the headers (and then physically destroy the disk). Don’t remember where I read about it, but been doing that for a while.

At least when it comes to Go code, I'm pretty impressed by the results so far. It's also pretty good at following directions, which is a problem I have with open source models, and seems to use or handle Claude's XML output very well.

Overall, especially seeing as I haven't paid a dime to use the API yet, I'm pretty impressed.