aws_ls

Joined 11 years ago

Karma
528
Submissions
308
Member since
Jul 1
As of September 2025: Since my contract ended sometime ago, I have been taking it easy. Now keeping an eye on work, which pays well. Until then, trying to keep the days fulfilling (i.e learning and building, instead of passive consumption). Its not easy, but some days I am good at it.

As of April/May 2022: A programmer who now mostly freelances at present. I am trying to get into big tech, just to keep the stuff interesting, and not getting too comfortable with a particular way of work mode.

I am a 25+ years experienced developer and software architect. With experience in various tech stacks. And always eager to learn new things (For example, was fascinated by Self-Driving cars and did the Udacity course to learn about the tech)

Up to 2018, I was trying to grow my own Startup. Dream of getting back to startup life, or my own projects, at some stage. Hopefully, when my kids have graduated college.

-- My note written, when I first created the account in 2015 --

A programmer who is trying to grow a Startup, organically. Like how a tree grows from a seed. Without the inorganic fertilizers aka VC money. Nothing against VCs. Just like the freedom for now, although could do better with the pace. No easy decision in life.

PS: Not affiliated with AWS. Just created this handle, when I had to ask an AWS specific question (my first post for this handle[1]). And then continued with it for other things.

[1] https://news.ycombinator.com/item?id=9810303

View on Hacker News

Recent submissions

Comment · 2 months ago
I code using Codex or sometimes ClaudeCode.

I use Zed to review the code.

CLI is great for most things. Some times, when I am reviewing a PR, I just tell Codex to do a diff between source and target branches in PR, and generate a HTML report. Then I read that, and also review relevant pieces of code in Zed.

I tried using Cursor, many of my colleagues use it, but stopped using it due to clunkiness. And of late, even in there, many of my colleagues use agentic coding. Instead, I prefer the snapiness of CLI (Codex coded in rust, takes less memory per session) and Zed (also coded in rust,) and blazingly fast. Zed is the new Sublime Text if you will.

Comment · 2 months ago
> “There was never a $BETR without @vishal_better. That demands respect.”

and

> “(Lewis’) thoughts about cost savings were good. His ideas about innovation were not,” Garg argued. “It’s so much easier when we’re this close for someone to come in and say that they could have done better.”

The user of better for Better, had me chuckling in these parts.

Overall, I think, we can safely say that Better has definitely seen some better times.

Comment · 2 months ago
> I feel like I owe it to the people who participated to keep it online forever. Also as a document of history.

Oh! So its your website. Cool.

Did the money change hands, for this bet, in 2022? How do you ensure it. Or is it honor based?

Comment · 4 months ago
Its surprising, how this key insight is lacking, even in the best of developers.

It happened fast, as you prompted it just the right way. Another person, who doesn't have all that context in the mind, would fail quickly.

For example I have decades of Software experience, but wont know where to even begin what OP did.

Comment · 5 months ago
Thanks for your explanation. I did not get it in the first read, and was too lazy to think, until saw your comment.

Just want to point out, that one is actually doing the experiment with a biased coin, then one must ignore all pairs.

e.g in case a coin which is heavily biased, say .9 H and .1 T. One should start with ignoring all the HH pairs, and start only at odd index. Lest, one picks a value like HHHHT (in the case the 2nd HH pair was not skipped, instead they greedily picked up the first HT, which will make the experiment HT biased).

Comment · 10 months ago
Yep. I too felt that. The insights seem genuine. But probably fell into the temptation to use LLM to structure. I feel increased cognitive load with numbered lists, thanks to LLM.
Comment · 11 months ago
Well its easy to cause damage by messing up the `rm` command, esp with `-fr` options. So don't take it as a proxy for some great skill which is required to cause damage.
Comment · 1 year ago
Yes, first thing, you must stop driving at crazy speeds. Good thing is is you already know what you need to do :)
Comment · 1 year ago
I am a curious sort. Always look at bios. Came across yours, and reading your blog. Keep it going! :)
Comment · 1 year ago
Go compiler tooklit works very well on all the OSes. That would be a consideration.
Comment · 1 year ago
Comment · 2 years ago
>What I mean is that when its generate a response, the computation happens on a snapshot from input to output, trying to map a set of tokens, into a set of tokens. Model doesn't operate on a context larger than the window

The weights in the model have the larger context, the context length size of data is just the input, which then gets multiplied by those weights, to get the output.

Comment · 2 years ago
For me google search is still useful for weather, stock prices, flight prices, maps, also to reliably navigate to a website(e.g. of a bank).

But for all other (code/health/taxes/so many others) queries use chatgpt. For code occasionally need to go to API docs, if chatgpt (v4) hallucinates. Not very often, but does happen, if the requirement gets complicated, example involve some specific (older) versions of certain APIs.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton