Hacker News: nikolay

New comment by nikolay in "I toured Unitree's factory. I'm more bullish on the West"

nikolay — Sat, 07 Nov 2026 10:48:13 +0000

Bullish or bullshitting?

New comment by nikolay in "YC's Garry Tan uses AI agents so much, he spends up to $200K a year on tokens"

nikolay — Fri, 06 Nov 2026 10:34:00 +0000

It looks like Matt Van Horn, who I've met when he was at the original Digg, also is contributing to all kinds of projects on GitHub and releasing his own. Coding agents could be a curse for a few, and a blessing for many.

New comment by nikolay in "Dutch 'idealist' bank ASN blocks non-Google Android, including GrapheneOS"

nikolay — Fri, 06 Nov 2026 10:28:32 +0000

Other European banks do it, too - N26 also rejects GrapheneOS.

New comment by nikolay in "Readwise Books: An independent digital-first bookstore and app"

nikolay — Mon, 02 Nov 2026 01:30:01 +0000

You announce the launch in advance so that people can prepare. Otherwise, you're discriminating people by time zones giving an unfair advantage to people who were awake when you announced it or weren't at work or didn't want to do personal stuff while at work. This is exactly how Facebook did it years ago - they didn't have usernames in the beginning, but announced at what date and time they will open it up to the world. But you didn't have to invest a new identity - you could've used Goodreads or X.com or another one common among your customers. Also, you reserve usernames in high demand and you create some automated granting process - the more social media accounts with that username you link, the higher weight in the claiming process you get, and so on.

But the best approach is not to launch a service with usernames - you gain NOTHING. In fact, you can only piss people off! You launch the service, you allow users to link other social identities and you then decide what to do.

New comment by nikolay in "Readwise Books: An independent digital-first bookstore and app"

nikolay — Sat, 31 Oct 2026 22:20:51 +0000

I am canceling my Readwise as you didn't get me an advance notice for username grabs and I couldn't get mine, so, I have no incentive to promote username I don't associate myself with. You guys need to understand the meaning and value of identity - you can't play games of rolling dice! Learn from how Meta did it with WhatsApp usernames - they did it right, and you screwed up!

New comment by nikolay in "Ask HN: How do you verify a reported domain sale is real?"

nikolay — Sat, 31 Oct 2026 20:33:49 +0000

What's the context?

New comment by nikolay in "Ejaculation frequency and prostate cancer (2022)"

nikolay — Sat, 31 Oct 2026 20:02:09 +0000

Anecdotal evidence: many monks end up having prostate cancer, including famous saints of Eastern Orthodoxy.

But didn't Tim Ferris in his bod say that the more often you ejaculate, the shorter you live? His rationale (based on talking to doctors, scientists, etc.) was that semen production is prioritized for the allocation of many nutrients. So if you do expend them too often, but don't get enough from your diet, you deplete your body of them.

New comment by nikolay in "Ask HN: Opus 5.5 suddenly dumb as rocks?"

nikolay — Sat, 31 Oct 2026 04:25:43 +0000

Yeah, I've tried this as well, but via Antigravity - it was very effective. I wish to have some smart router that monitors the trends of your usage and has skills-based routing to the best model for the job.

By usage trend I mean something considering Codex almost predictable resets, your usage, and maybe even plan ahead using your backlog to know what you will need sooner.

For example, if you want to ship a lot of UI/UX this week, save Claude for it, and use other models to do the coding.

In fact, I completely revamped the way I do testing and even GitHub Actions, because I started to get hit with a lot of GitHub Actions costs. So, what I did was a smart test actions, which verify if the code has test attestations for local runs, so, they just skip burning GitHub Actions redundantly - the code has already been tested successfully locally.

Another huge burn from Actions came from squashed PRs - my attestations didn't work as the commit hash changes, so, I needed to put that into consideration as well.

In general, I'm completely revamping my release cycle, not allowing PRs to age, and main no longer represents Production. Agents keep merging as soon as possible and then another process assesses if there's a good enough release candidate having enough features that pass more extensive tests. There's no point to always test and always fix piece by piece if you're not going to release.

So, the RC candidate also generate a changelog, blog posts with the features, and so on.

Laws You Can Feel – Laws and heuristics deciding how an interface feels to use

nikolay — Fri, 30 Oct 2026 21:43:56 +0000

Article URL: https://lawsyoucanfeel.com/

Comments URL: https://news.ycombinator.com/item?id=50310548

Points: 2

# Comments: 0

New comment by nikolay in "Ask HN: Opus 5.5 suddenly dumb as rocks?"

nikolay — Fri, 30 Oct 2026 06:03:27 +0000

Yes, it's about the same time when I noticed degradation in the quality, too.

Elephant-Goldfish Model

nikolay — Fri, 30 Oct 2026 03:36:50 +0000

Article URL: https://egm.dev/

Comments URL: https://news.ycombinator.com/item?id=50299600

Points: 1

# Comments: 0

New comment by nikolay in "Ask HN: Opus 5.5 suddenly dumb as rocks?"

nikolay — Fri, 30 Oct 2026 00:41:58 +0000

It is dumb. None of my threads were closed with shipped features. It keeps going into endless review cycles essentially burning all my quota within a couple of days. And I'm not talking about complex a features an OpenAPI-based integration and some enhancement of pytest traces building maps of touched files. Both these are in their second week of token burn!

New comment by nikolay in "Ask HN: Opus 5.5 suddenly dumb as rocks?"

nikolay — Thu, 29 Oct 2026 21:48:44 +0000

Agreed. Yet, it munches tokens like there's not tomorrow. I don't get anything done with Claude in the past 2-3 weeks - just eating up its weekly quota within a day.

New comment by nikolay in "DoorDash to acquire Instacart"

nikolay — Thu, 29 Oct 2026 20:43:40 +0000

I use both for grocery delivery. The grocery platform of Instacart is better, so, I hope we can get the best of both worlds with having the better DoorDash promos with the better platform.

New comment by nikolay in "MacBook Pro with PROLED XDR Display and Touch"

nikolay — Thu, 29 Oct 2026 18:54:45 +0000

No M6, really? Anyway, I've never used touch on Windows, never going to is it on macOS either.

New comment by nikolay in "Claude Code regularly sends each background agent’s transcript to its model"

nikolay — Thu, 29 Oct 2026 18:29:41 +0000

Running in parallel and the new "thread coordination" is even worse! I "love" the cryptic messages threads share, and the endless loops of "main has changed, so, all tests done got invalidated". Maybe I have the wrong workflow, but I doubt. Mostly it's harness regular updates.

New comment by nikolay in "Claude Code regularly sends each background agent’s transcript to its model"

nikolay — Thu, 29 Oct 2026 09:49:41 +0000

A domain names web storefront with negotiations system, complex payment shapes, quite a few integrations with registrars and services, a Telegram bot for back office with mini apps, etc. A CLI, a TUI, an MCP, a WebMCP, text messaging. Built mostly on Cloudflare and with Ansible and OpenTofu. Runs also a local DuckDB database as I don't feel comfortable putting registry keys somewhere in the cloud, and so on. It is a scope creep, no doubt, but the idea is a to be fully autonomous, to ship features after I just pitch an idea to it. It writes blog posts, produces podcasts (not using Gemini Notebook!), advertises on social media, etc. Lots of work I could never do without coding agents, and I try to perfect it from all angles - following standards, and being as modern as possible. It's overkill, no doubt, but hey, why not. I'm trying to build a foundation for future projects, and get back into real development, and the last 10-15 I spent mostly with Infrastructure as Code. The maintenance of the entire system is probably less than $50/mo, which includes token costs for LLMs, ElevenLabs, etc.

New comment by nikolay in "Claude Code regularly sends each background agent’s transcript to its model"

nikolay — Thu, 29 Oct 2026 09:32:30 +0000

I constantly get into endless review cycles of 50+ rounds. Claude Code simply cannot get many things right regardless of the effort.

New comment by nikolay in "Claude Code regularly sends each background agent’s transcript to its model"

nikolay — Thu, 29 Oct 2026 09:31:02 +0000

I am talking about functional flaws, not cosmetic ones. For example, in web apps, where a11y is mandated, Claude Code kept violating accessibility (regardless of the installed skills), and even the lackluster GitHub Copilot with some outdated models always finds serious violations. Codex constantly finds race conditions and other timing flaws for distributed services written by Claude Code. But I have to admit that Codex sucks at web design compared to Claude Code!

New comment by nikolay in "Claude Code regularly sends each background agent’s transcript to its model"

nikolay — Thu, 29 Oct 2026 06:01:16 +0000

Opus 5.5 burns tokens like there's no tomorrow! Meanwhile, GPT 6.1 Sol on Extra High feels unlimited! Opus (on Max 20x) munches my weekly quota in a day, or two at best. Meanwhile, working with Codex and reviewing with a $50/mo Muse Code Wonder Full seems like it can go own forever (on the same codebase, the same skills, etc.) I am highly disappointed by the overly glorified Opus and Claude Code in general - there's no commit in which Codex or Muse Code or GitHub Pilot don't find critical flaws! I have long sessions of Claude Code having over 100 review cycles of Codex and still not addressing all the valid issues that Codex finds in Claude Code's work! Claude Code keeps injecting British English into prose and identifiers with nothing being able to stop it! Of course, there are almost never resets for Claude Code either - they would rather me focus on Codex being my primary and Claude Code being the reviewer (it works much better than in reverse). Anyway, enver trust a single model! Always review with another frontier one!