malkosta

yes true, but I already have it on all my machines, for diverse reasons...it wouldn't be hard to implement a markdown support, but it would be worse to maintain that

mine is:

find . -name '*.md' -type f -exec sh -c '

  for file do  
    out="docs/${file#./}"  
    out="${out%.md}.html"
    mkdir -p "$(dirname "$out")"  
    pandoc --quiet --template template.html "$file" -o "$out"  
  done  

All I remember is that it's pretty clear written in the post...

The problem with QWEN is that it just can't edit files reliably, I had to hack Pi all over to reduce the pain, but still far from perfect...does Gemma 4 strugle on this?

That was exactly my same question. Then I finished reading the post. The reason is pretty clear, and written in the post: it is faster than ollama+mlx.

What about just getting work done because you like to code, and not giving a damn about what others think because we will get rich anyways…I wonder why people are always trying to win instead of just having fun.

What a master class in linear algebra…

Yes, it describes human nature better than psychology. We can’t fight even knowing about it.

How do you ensure security? Does it have a proper container? Otherwise, it's impossible to prevent leaking prod secrets.

How do you intend to conquer the preemptive scheduler? System-wide fairness and preventing starvation are essencial steps for this to work well.

It’s crazy how pretty much every tool people post to support AI systems is already in Erlang/OTP or in elixir standard libraries.

Elixir + Postgres is the microservices killer...last time I saw VP try to convince a company with this stack to go microservices he was out in less than 6mo

You read my mind sir! Also, great work by OP, I really needed this since a long time!

Beautiful! :clap:

Rectangle+Apptivate made me stop looking for an i3 alternative, after years. The first for moving windows, the second for switching between them with super+number, just like i3.

I fight against this by using it mostly on trivial tasks, which require no comprehension at all, also fixing docs and extending tests. It helps me to focus on what I love, and let the boring stuff automated.

For complex tasks, I use it just to help me plan or build a draft (and hacky) pull request, to explore options. Then I rewrite it myself, again leaving the best part to myself.

LLMs made writing code even more fun than it was before, to me. I guess the outcomes only depends on the user. At this point, it's clear that all my peers that can't have fun with it are using it as they use ChatGPT, just throwing a prompt, hoping for the best, and then getting frustrated.

I do almost the same, with a simple bash script, and no JS junk: https://github.com/alexandremcosta/alexandremcosta.github.io...

TODO is a pretty clear name. If it's not to be done, then it should be just a regular comment, no need to tag as TODO.

If it's not intended to be done, then you are using the wrong name. Don't call it `TODO`.

It's just a ~~reduce~~ loop, with an API call to an LLM in the middle, and a data-structure to save the conversation messages and append them in next iterations of the loop. If you wanna get fancy, you can add other API calls, or access to your filesystem. Nothing to go crazy about...

Offset-based pagination will be a problem on big tables.

The official ones are still the best: https://hexdocs.pm/phoenix/overview.html and https://hexdocs.pm/elixir/introduction.html

I have a 60 LOC [bash script][1] to do something similar to my personal blog

[1] https://github.com/alexandremcosta/alexandremcosta.github.io...

Also [Homerow](https://www.homerow.app/) is really good

Too bad you didn't choose Phoenix/Elixir. Give it a try, and never come back to Rails.

Just a simple webapp I made to help with decisions some of us have to make every month.

It reminds me of Joe Armstrong's "My favorite Erlang Program": https://joearms.github.io/published/2013-11-21-My-favorite-e...

To focus on short-term speed instead of long-term speed.

Amount of 4xx, 500, and 2xx of an http application can tell a lot about application anomalies. Other protocols also have their error responses.

I also keep a close eye in the throughput VS response time ratio, specially the 95th percentile of the resp time.

It’s also great to have this same ratio measurement for the DBs you might use.

Those are my go to daily metrics, the rest can be zoomed in their own dashboards after I first check this.

I have a friend who is not tech savvy at all. And I need him to send me a huge file. Is there any GUI frontend for this?