I really like cmux for this (https://github.com/manaflow-ai/cmux)
HN user
funerr
https://unzip.dev
ChatGPT summary:
Prototaxites was a massive, trunk-like organism (up to ~8m tall, ~1m wide) that dominated land ~420–370 million years ago, long before trees or complex plants existed. It looked like a tree, but chemical evidence suggests it didn’t photosynthesize. Internally it was made of interwoven microscopic tubes, unlike plant tissue. It’s often described as a giant fungus, but it doesn’t cleanly match modern fungi either, and some researchers think it may represent an entirely extinct branch of eukaryotic life. In other words, early “forests” may have been dominated by something we don’t have a modern analog for.
ai-sdk by vercel?
When you have LLM requests you don't mind waiting for (up to 24h) then you can save 50% in costs. Great for document processing, image classification at scale, anything that you don't need an immediate result from the LLM provider and costs play a role.
I needed a Python library to handle complex batch requests to LLMs (Anthropic & OpenAI) and couldn't find a good one - so I built one.
Batch requests take up to 24h but cut costs by ~50%. Features include structured outputs, automatic cost tracking, state resume after interruptions, and citation support (Anthropic only for now).
It's open-source, feedback/contributions welcome!
How does this compare to better-auth?
I think it is actually a solid choice given the startup ecosystem and generally easy async nature.
I met the founders, great people. Any open-source tools to start out dealing with failed payments before we scale?
I didn't know that about Maybe Finance. Was that what prompted you to os inboxzero?
What are the pros and cons? Aren't you afraid that it makes the SaaS offering less enticing, especially to highly technical people (which seems like might be a big chunk of your potential users)?
Retool for sure, they are pumping out features really quickly too. The model makes a lot of sense from a developer's point of view. Now that they are also going into public apps I would really consider them seriously.
Hoping for: - Better pricing for public apps. - Better ability to easily customize the frontend design without touching CSS/HTML/JS.
There are too many variables to predict the future. Yet, I believe that it will empower more 1-3 person businesses to compete with the big incumbents.
Imagine you could run teams of marketing/sales/support ai-agents and focus on your core business. "The next 1B, 1 Person companies".
Just 15 years ago you'd need teams of 10+ people to do basic ml tasks you could do today. Now you have an LLM or a SaaS service that outperforms those teams for the fraction of the cost.
I just migrated from Ghost because many of these problems, and lack of basic features like an integrated referral system for my newsletter.
I'm glad you like it.
Shameless plug, I just wrote about Vector Databases and Embeddings 2 weeks ago in my newsletter (5 minute dev trends):
https://unzip.dev/0x014-vector-databases/
It also explains how to use them IRL and why.
My thoughts exactly.
Cool idea! I thought about it too a few years ago while doing some upgrades. Do you support more languages? And do you have it in CLI form?
I personally use a bit of a different prompt, so it's tailored for my repo. But yeah, you can also customize it pretty simply, like you mentioned in the ./src/ai.ts file.
This is a blueprint on how to have checks in the CI, it took me sometime to build and get working - do you know anything that does this out of the box? Thanks for the feedback eitherway!
How is this different from chatbase.co ? (not affiliated)
What did you exactly do to find it?
Thanks tomgs! I have this format for every new dev trend I find, keeps it structured.
Happy to help!
Shameless plug, I just finished 30+ hours of research into io_uring and posted a more high-level summary here: https://unzip.dev/0x013-io_uring/
Including a tl;dr, pros, cons, tools & players, and a forecast.
The creator of io_uring Jens was super nice and also reviewed it (I'm still a bit in shock).
Thanks for the shout-out!<3
Sounds really similar to Property based testing [1]. (Rule-based testing)
Author here, I wrote at the begining of the text that it isn't fully supported yet :)
"WebGPU was first published in mid-2021 (super new!), so it isn’t fully supported on browsers (also shows how you can enable it) yet".
My main goal is to bring this new technology into attention, I don't intend nor do I have any incentive to recommend people using it in production. Sorry if that wasn't completely clear.
Added! thank you!
I'm the author, not the poster. Unzip is more of an intro summary of a concept so the title is indeed a bit misleading ;)
Thanks for including Unzip nonetheless! Yeah, it's more of an intro :)
Thanks for the recommendations, adding to the extra section!
Hi HN, I'm the author, but not the poster, I didn't plan on posting here yet.
Ps Unzip is a summary/intro not an "in indepth article". I try to summarize concepts. But thanks to whoever posted, appreciated either way.
AMA