softwaredoug

I still think I haven’t gotten much more productive than a year ago: I wrote the barebones, scaffolding, and where I cared about the details. Agents stamped out the rest from those patterns. Everyone …

One day we'll get to the plateau of productivity of this hype cycle and wonder WTF were we thinking. Not that we won't have most code written by agents. I truly beleive that to be true. But …

But why aren't lower level "ideas about execution" more clearly expressed in code? Or at least pseudocode? Many of us would struggle with our reading comprehension of an English descrip…

This article feels too black and white > How are you supposed to review 5k lines of code every day? Maybe there’s grey areas of reviewing 10% of the code, ensuring tests actual are meaningful, the …

I write 2% of code. And read 5-10%. I feel like everyone is being too black and white. And everyone has different ways of staying in flow with a software task.

Probably because agents in typescript don't waste time reasoning about so many different problems that come up in a lower level language. They stay higher level, and don't waste context on t…

Exactly this. Spending a few hours crafting and debugging a test you trust save hours of pointless agent time.

There are ways humans have an advantage. Not in the line by line code, but in the experience of having the whole mental model in your head. Writing some code with the whole context in your head is the…

I personally am advocating for writing small amounts of code because I think it makes you more productive Spending a day manually reworking the architecture in my experience can save you $1000s in tok…

What I mean by "talk to the agent in code" really is to code. Establish patterns. Then have the agent use those patterns as reference for other areas of the project.

Do you like to look at code? I’m finding it’s really just a personal preference. And honestly not worth getting worked up about either way.

My other article on this topic advocates for using code over specs :) It’s ok to talk to the agent in code. Or create examples for it to follow. https://softwaredoug.com/blog/2026…

I think what we’re learning is that up front slower experience can be valuable for maintaining the maintainers own understanding of the project I’ll be very curious how / if the Bun port to Rust …

I’m finding it all boils down to cognitive differences. Some people find code easier to read than the English description. It’s more precise and many experienced devs can scan it and know what’s happe…

I’m old enough to remember coders being made fun of for doing copy-pasting from stackoverflow all day.

The thing is with a wish machine, you can one shot something very easily. But then you realize software is the accumulation of 1000s of wishes. And you want this but not that. Many little micro decisi…

How do you arrive at a failure case to measure without looking at the code? How do you establish sane patterns when you’re in an exploratory /architectural mode? You could do this in English, sur…

Yes but I always have to be on the lookout for this meta pattern that leads to code bloat. In this case things mostly sorta worked and the simplest way to see the problem was look at the code. And try…

Yet as a C developer for 15 years I frequently looked at asm and on occasion even wrote a little. And that is a far stronger abstraction than LLMs :)

I was fairly confused myself as the author :) This was actually my original submission last week. There was a front page submission last night from someone else (hence the comments). Then my old post …