Hacker Newsnew | past | comments | ask | show | jobs | submit | cejast's commentslogin

> The apps I build tend to be the result of several dozen or several hundred features developed mostly serially, and I am always the bottleneck because of that whole "knowing what I want" thing.

This is probably where the disconnect is - the TUI workflow only really clicked when I started working across multiple worktrees at the same time.

It’s not all or nothing. You can still be fully in the loop, read all the code and steer each session towards the outcomes you want. You’re just doing it across multiple branches at the same time and imo it’s easier to do that through the TUI.


How do you stay effective when working across multiple worktrees at the same time? That is, how do you keep context in your head?

Then when it's all over, how do you evaluate which one to keep?


This is indeed a challenge and there is a practical limit to how many you can manage at once. It really depends on what I’m working on, for a sufficiently large enough project there are lots of independent tasks that can be delivered in isolation and so it’s rarely a case of babysitting all of them in realtime.

I tend to have one session that I’m actively focused on as my primary session, everything else I come back to every ~5-10 minutes to nudge them along while I’m waiting on the ‘main thread’. If my core task demands more of my time to think then I check less often.

The important thing is to find what works for you. If Cursor is giving you enough productivity then that’s awesome, carry on with it! I would recommend trying out other workflows, weighing up the pros/cons yourself and ultimately find what works for you. TUI-driven development isn’t for everyone in the same way not everyone likes to use vim and that’s okay.


I appreciate your conclusions, and frankly, you're probably right.

I'm still quite curious how folks evaluate between four completely separately but presumably functionally similar outputs and conclude that work tree 2 is the winner.


I’d say it’s important to understand how they work but I also rarely find myself reaching for channels. I see more usage of wait groups and mutexes, but even then you can build abstractions around these in a way that can be reused without having to touch them again.


What kind of patterns?

I've started using golang last year and I feel like I'm missing exactly this kind of experience with these patterns


I have to regularly travel to rural villages in Asia and the people that live in these places have only the most basic human desires. They barely know what an LLM is let alone interact with them. They're some of the happiest people I've ever met.

Of course they'll complain about work or money, and they assume that a Western salary buys me freedom. But they get to live through these luxuries on a daily basis. I'd be lucky to enjoy just one of these in any given week.


Not to be glib, but I assume you're not quitting your Western job and moving there. So isn't this romanticizing it?


Yeah as someone whose family left this, you guys are crazy


I have my own family dependencies that aren't too keen on the idea of living the village life. Otherwise I'd already be there.


I saw a video of a guy in his 50s living in a small city in Vietnam. He claimed $800/mo was more than enough, and $1000/mo would let him live lavishly. He just did the math and as soon as his predicted months of life left * 1000 < savings he pulled the trigger.

I think about that a lot because I have already hit that number easily, my family would definitely suffer though and not be into it.


There is a reason that South East Asia is popular for digital nomads and for retired expats. Thailand and Vietnam have still very affordable places if you avoid tourist traps. I would recommend first trying some easier destination like chiang mai or danang (if first time in asia), staying for 2 months (vacation/workation) and seeing if you still like it.


It’s not that easy to emigrate afaik, even ignoring family and obligations like that.


for real happiness it would be the most important thing if not romanticizing.


But is that true though? This seems bit of a romanticism. Having lived in rural villages in Asia, I can tell you people are about as happy as anywhere else. They worry about not being able to afford or have access to basic healthcare. They worry about one off season rain destroying their livelihoods for the season. Depending on where you are, some worry about a sudden onset of escalating violence. Plenty of people don’t even have access to drinking water. Some worry about their land and only source of income taken away by larger interests.


Desires only tend to magnify in intensity and multiply in numbers because there is rarely enough.

Working on desire is a big thing.


Welcome to the AGI era - where human intelligence is our only constraint


For what it’s worth, I find comments and articles with assertive predictions like this difficult to take at face value.

I don’t even disagree with the premise, but it shifts the burden of assessing likelihood back onto the reader, so it doesn’t really add much value to me.


> All my finance and payment domain expertise, all the debugging intuition and distributed system knowledge earned through hours of sweat and tears, is now promptable.

Is it really though? Access to information is quicker, but you still need to know what ‘good’ looks like to leverage it effectively. I can prompt my way to a medical diagnosis, but I’d still want to run it by a doctor.


I’ve found it extremely hard to get LLMs to exit the basin of your current knowledge.

One of my tests for new models is to ask about a concept I already know the mathematical model for, but as if I don’t. So far, they all answer the same way:

1. Convoluted explanations about how it kinda-sorta is common terms.

2. If you follow up with the correct mathematical term, it immediately claims that’s correct and the right way to model it.

3. If you ask it why it didn’t use that term for your question, the LLM gives some version of explaining that it tried to match your language.

I have no choice but to assume the model behaves similarly other times — and that I am largely trapped in a basin of my own ignorance, when using LLMs.


I don't think that's a good analysis.

If the LLM is wrong and gives you a wrong medical diagnosis you end up hurting your health. If an LLM gives you a wrong debugging answer you've just lost 5 minutes.

Software engineering is the only knowledge work where mistakes are usually inexpensive except for data breaches. Outside for that nobody cares for bugs.

That's not true in most other knowledge jobs. If a lawyer uses AI and hallucinates something there is a legal problem. If someone vibecodes an app and crashes, it can be fixed with more AI and try again


That’s my point though? Debugging a 5-minute problem is in the shallow end of the spectrum, the real complexity sits where they lean on their domain experience. Finance and payments software mistakes can absolutely be expensive.


> More recently, it's "this failed because Claude chose..." No, Claude didn't choose, the person who submitted the PR chose to accept it.

I can relate to this, unfortunately these tools are becoming a very convenient way to offload any kind of responsibility when something goes wrong.


Well if the management want to get more AI, they gonna get more AI, and no, I am not gonna be running around making sure their dreams work smoothly under my human supervision - I am gonna let it go all the way they want. In the mean time I focus on improving my skills.


This was my reaction as well, a lot of hand-waving and invented jargon reminiscent of the web3 era - which is a shame, because I'd really like to understand what they've actually done in more detail.


What's even more interesting is that if you turn off watch history, they disable the home feed altogether. They just give up trying to show you anything, which has been great for keeping me off YouTube.


That is a very broad generalization. Even if it was 'put into some fund', that equates to a capital investment which can be used to deliver value elsewhere.

Money is complicated - the only way in which I would see it get truly wasted is if you took it out as cash and burnt it. Even then you'll be (marginally) raising the value of all other money left in the system.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: