HN user

shanev

894 karma

http://twitter.com/shan3v

[ my public key: https://keybase.io/blockshane; my proof: https://keybase.io/blockshane/sigs/t8BAIrc2bPHKQJ_46pr9XRUWVKywSj0RAjDpfAfL9Ko ]

Posts11
Comments215
View on HN
GPT-5.2-Codex 7 months ago

The GPT models, in my experience, have been much better for backend than the Claude models. They're much slower, but produce logic that is more clear, and code that is more maintainable. A pattern I use is, setup a Github issue with Claude plan mode, then have Codex execute it. Then come back to Claude to run custom code review plugins. Then, of course review it with my own eyes before merging the PR.

My only gripe is I wish they'd publish Codex CLI updates to homebrew the same time as npm :)

This is solvable at the level of an individual developer. Write your own benchmark for code problems that you've solved. Verify tests pass and that it satisfies your metrics like tok/s and TTFT. Create a harness that works with API keys or local models (if you're going that route).

DAOs are fixing this in the crypto world. You contribute to the protocol and get paid by the DAO. Everything is transparent and open. If you do this enough and earn the respect of the dev team it could even turn into a full-time role.

Brad Cox has died 5 years ago

He worked hard to enable software reuse. No one was interested in his idea of trying to monitor component use during runtime to pay developers

People are experimenting doing this in blockchain smart contracts. It’s transparent and supports micropayments as well.

Khan Academy’s economics and finance videos are excellent for learning the fundamentals.

Ray Dalio’s talks on YouTube, and his most recent TED talk are great for context around economic downturns.

Real Vision and Anthony Pompliano’s podcast are good for deep dives into what’s going on with a bit of an alternative take.

_why’s contributions to the Ruby world always had an art project element to them. His disappearance is just an extension of his art.

It was amusing to me that his real name wasn’t even exposed on conference badges.

When a company switches languages like this, it's usually because the engineers want to learn something new on the VC's dime. They'll make any excuse to do it. As many comments here show, there are other ways to solve this problem.

This is a common pattern with many of Go’s features. Experienced folk sometimes tend to make unconventional design decisions. You don’t appreciate them until you’ve tried out all the alternatives, and arrived at the same conclusions yourself.

Since the App Store launched in July 2008 until Nov 2011, my app, Drum Kit, was one of the top apps in the Music category. It went up to #2 at one point, only behind Smule's Ocarina, and had over 10 million downloads.

The day GarageBand launched, my app went from something like #24 in Music to #64, and downward from there. Fun times!

Interestingly, at an Apple conference sometime in 2010 (either C4 or WWDC), some guy asked me all kinds of detailed questions about the custom audio engine I wrote. Back then the "iPhone SDK" (now iOS SDK) audio APIs had too much latency for a drum app. So I had to write my own in C++ based on Audio Units (a low level C API). A tricky thing I had to do was mix multiple buffers of audio to create the resonance effect drums make when you hit them repeatedly (most prominent in cymbals). Later I found out the person asking me the questions worked at Apple, on the team that made GarageBand. They didn't launch with a blended resonance effect. When you hit a cymbal repeatedly, it would cut off the sound and play the sound again. Ew.

People miss out when they don't try some of the more indie apps. A lot of care goes into making them.

We have a couple of mechanisms in place for this.

Users who have earned a certain amount of the token by being good citizens of the network will have to vote in new users. These users also moderate content, where offenders get some of their token taken away. Furthermore, network operators (validators) can vote to remove malicious content and users. And as with most crypto networks, a small fee (gas) is charged for usage.

This is what we are aiming for in building TruStory (https://beta.trustory.io).

The protocol is built as a distributed proof-of-stake network, where people can earn a crypto token by contributing good content, or by running nodes of the blockchain. Users and node operators can take ownership in, and govern the network based on how much of the crypto token they own (or have earned).

I recently wrote about the architecture at https://blog.cosmos.network/building-a-decentralized-app-wit....