RamtinJ95

who said they are unable to work? This meme is so old and overused.

I get calling opencode a hack, but actually looking at the product claude code is the hack compared to opencode for real

Wow did not expect to get a thank you from the legend himself! Thank you for your great courses and the care and effort you put into them sir! :D

if you are looking for a resource for this, I did this exact thing last year through pikuma that man is the best software teacher online I have come across. Highly recommend his 3D software renderer course.

This article makes sense it really does, but its not the full picture. I think there are different modalities to enjoying programming. I wrote a long post about this a couple of months ago that goes way more into detail than I could ever write up in a HN comment. article: https://handmadeoasis.com/ai-and-software-engineering-the-co...

Awesome thank you looks like a great intro to the idea at least

This sounds very intriguing, any resources around this approach you can point me towards?

Pikuma.com writes a software renderer pretty much from scratch with all the necessary math and explanations in a very pedagogical way. Highly recommend it

I concur with the other commenters, 4.5 is a clear improvement over 4.

As a counter point to this. I have been using opencode for months and it has been stable for me. Also on linux and tried it on alacritty, ghostty and kitty works without a problem in all of them. To me its as good as claude code but i have spent some time tinkering with it and set stuff up the way i like and developed some plugins i needed for it also.

Work on side projects or plan the next step with another agent is what I usually do. For example I have been learning and coding in golang lately while llm does some grunt work for work tickets, I love it

This is exactly what I experienced as well a few weeks back and wrote about here: https://handmadeoasis.com/ai-and-software-engineering-the-co...

Essentially it boils down to what kind of programmer you are and what about the craft of software engineering you enjoy.

Just use gh_grep mcp and the model will fetch what it needs if you tell it to, no need to download from GitHub manually like this

This looks really cool and I love the idea but I will stick with opencode run ”query” and for specific agents which have specific models, I can just configure that also in an agent.md then add opencode run ”query” -agent quick

Something like opencode probably, that’s what I have been using to freely and very easily switch between models and keep all my same workflows. It’s phenomenal really

I hate python and I use it everyday because I work in the data space. Its a toy scripting / glue language that has gotten used for far too much that it was not designed for. The usual suspects are also really annoying, such as white space instead of {}, no types, its so damn slow and all projects use so many packages and the packages use packages etc. That last one could just be a personal preference thing to I will admit, but the rest are just almost objectively bad. Especially when building infrastructure like a data platform.

At that scale it cannot be cheaper than just running the same workload on BigQuery or Snowflake or?

That is actually a great idea, and I agree it would be very useful to have such a catalog/map!

I wonder though if the space is mature enough for such a map or if it would become to generic to say anything meaningful.

I fully agree with that statement, but I dont agree with the premise that, that is whats happening currently.

I think "Identity Crisis" is a bit over dramatic, but I for the most part agree with the sentiment. I have written something in the same vane, but still different enough that I would love to comment it but its just way more efficient to point to my post. I hope that is OK: https://handmadeoasis.com/ai-and-software-engineering-the-co...

what makes it more likely that claude would read these .md files then?

Do you have any links to sources around these other ways to get the data? I recently started working in this domain and trying to get up to speed on state of the art regardings these types of things specifically since im in the data team.

This analogy is inaccurate and spreading it just plays into the hands of those trying to press down SWE salaries and their importance. A better analogy is that previously we had a hand saw, now we have one of those automatic ones which speeds up some processes a lot but cant be used everywhere in every situation.

This is a topic I feel very strongly about and have given it some considerable amount of thought. Most of those thoughts ended up in this blog post: https://handmadeoasis.com/ai-and-software-engineering-the-co...

But the whole blog is a consequence of exactly this that you are describing.

I don’t get it. I have been doing something similar for a month with opencode, is the new thing that the new sonnet model is fine tuned to call these tools “better” or simply that they have improved the devex to accomplish these things?

The db example is bad because most people in db and data world agree mongodb is not a great choice anymore. In fact mongodb today has actually adopted a lot of the characteristics of a sql db.

What? Who has made this claim, what is the evidence? I don’t think this is true at all.

This is not my experience at all I have been daily driving arch for months for work and hobby development without a single issue. In fact it’s easier to fix issues or keep things like my config working on arch than it was on macOS.

Yes in terms of raw LLM, but with some tools or a MCP the “AI” will never be wrong.

Linear is what you are looking for