HN user

gregwebs

4,424 karma

Software engineer.

Interests in security, startups, health, sustainability, productivity, community.

Side projects:

* eatnutrients.com * docjumper.com

github: gregwebs

gregwebs.at.hn

Posts56
Comments1,366
View on HN
www.warp.dev 1y ago

Bringing the Warp Terminal to Windows: Eng Learnings (So Far)

gregwebs
3pts1
www.consumerlab.com 3y ago

How to Avoid Buying Counterfeit Vitamins and Supplements

gregwebs
1pts1
reasonstobecheerful.world 3y ago

Heat from an Amazon Data Center Is Warming Dublin’s Buildings

gregwebs
2pts0
www.tabletmag.com 4y ago

How the CDC Abandoned Science

gregwebs
21pts3
www.lynalden.com 5y ago

How Market Capitalization Works (and a Look at Rolling Bubbles)

gregwebs
2pts0
www.epbmacroresearch.com 5y ago

Fiscal Stimulus: The Economy's Third Strike

gregwebs
1pts0
www.ft.com 5y ago

Chinese data surveilance is a messy beauracratic process

gregwebs
1pts0
www.pnas.org 5y ago

Antioxidants prevent health-promoting effects of physical exercise (2009) [pdf]

gregwebs
116pts39
www.sciencedirect.com 5y ago

Vitamin D improves survival in Covid-19 elderly patients

gregwebs
274pts107
medium.com 6y ago

Lab-Made? SARS-CoV-2 Genealogy Through the Lens of Gain-of-Function Research

gregwebs
2pts0
medium.com 6y ago

Coronavirus: How to Do Testing and Contact Tracing

gregwebs
1pts0
medium.com 6y ago

Coronavirus: The Basic Dance Steps Everybody Can Follow

gregwebs
1pts0
www.medrxiv.org 6y ago

DIY face masks with nylon stocking on top can work better than surgical masks

gregwebs
3pts1
aws.amazon.com 6y ago

AWS announces offline mobile DB that syncs to dynamodb

gregwebs
10pts0
blog.twitter.com 6y ago

Rezolus: A high-resolution systems performance telemetry agent

gregwebs
28pts7
nats-io.github.io 6y ago

NATS (similar to Kafka) 2.0 released with secure sharing and WAN deploy

gregwebs
3pts0
medium.com 7y ago

A Redis layer built on top of TiKV, the distributed Key-Value store from TiDB

gregwebs
5pts0
techcrunch.com 7y ago

“Commons Clause” was intended to stop AWS from selling open source as products

gregwebs
43pts40
www.hashicorp.com 7y ago

Announcing HashiCorp Consul and Kubernetes

gregwebs
3pts0
thenewstack.io 7y ago

Questions for Evaluating a Distributed Database

gregwebs
3pts0
www.pingcap.com 8y ago

Ele.me serves food to millions with Redis and MySQL on top of TiDB

gregwebs
4pts0
www.percona.com 8y ago

TiSpark: Data Insights from TiDB, No ETL Required

gregwebs
5pts0
vimeo.com 8y ago

“self-driving” databases (ML as DBA)

gregwebs
1pts1
github.com 8y ago

MySQL GraalVM Multi-Lingual Environment Demo

gregwebs
2pts0
pingcap.com 8y ago

TiDB 2.0 Is Ready – Faster, Smarter, and Battle-Tested

gregwebs
58pts10
blogs.oracle.com 9y ago

Micro-containers: rationale and tooling

gregwebs
3pts0
flynn.io 10y ago

Stable release of Flynn – open-source container deployment

gregwebs
79pts22
www.cell.com 10y ago

Modern Hunter Gatherers sleep 6-7 hours per night

gregwebs
1pts1
foundationdb.com 11y ago

FoundationDB Table Groups pre-join tables

gregwebs
1pts0
www.youtube.com 11y ago

Shen: A Sufficiently Advanced Lisp (Strange Loop Conf)

gregwebs
3pts0

MikroTik is one of the only companies that sells a router without WiFi at a non enterprise price. Useful for me to completely turn the power off at night to a separate WiFi (router in Bridge mode).

GPT-5.6 13 days ago

I run my AI agent as a different user (in addition to using the sandbox functionality provided by cc/codex). It does not seem possible to run the Codex GUI as a different user. I can run the TUI (/Applications/Codex.app/Contents/Resources/codex) but it has the shortcoming that remote control is only available in the GUI.

I installed the Claude Code Codex skill provided by Anthropic and I am having Claude invoke it automatically to review all plans and changes. The nice thing about this is that for an additional $20/month pro plan I can extend the runway for Claude rate limiting and compare frontier model responses. I am looking for more ways now to work in Codex as a subagent that gets used automatically from Claude Code.

No code review by default goes against actual established evidence (there is little of this for software development practice) that code review is the best way to find defects.

I always get the impression from using hardware and other anecdotes like this that it is rare for hardware companies to know how to do software development well because their core competency is hardware. In fairness, it is uncommon for software companies to know how to do software development well.

Every form of testing has its value when done well and they are using several forms that most software developers don't use- probably helps make up for the lack of code review and unit tests. But if they incorporated code review and unit tests their software would likely be even higher quality.

Property based testing is amazing, but it won't provide full coverage. Regression suites are amazing, but generally the most expensive form of testing in terms of time to write and maintain tests and time to run them.

Today AI can crank out unit tests so its silly not to have them.

All these conversations seem like they are missing talking about planning vs execution. I want the best possible frontier model to plan out my changes. I also have a 2nd agent that is a frontier model check the plan. Then at that point the implementation can be done by a lesser and possibly local model. The frontier model can still do a final code review on the implementation of the changes.

Claude code supports this by setting the model to "opusplan"- it will automatically use Opus for planning and sonnet for implementation. This was completely necessary with the fable release. I was able to do this with fable and it was necessary to avoid getting quickly rate limited. In settings.json:

"env": { "ANTHROPIC_DEFAULT_OPUS_MODEL": "claude-fable-5" },

Obviously have that set to "claude-opus-4-8" now.

Samurai City 2 months ago

This is fascinating after learning about and rading Peter Turchin's concept of Eelite overproduction [1]. The theory is that much of society's conflicts are actually fights between the elite and that this happens when the elite start fighting for the same resources.

From that point of view they seemed to have created a system that stopped the elites from starting wars with each other by imprisoning their families. And although they levered high taxes they did force many elites to accept a small amount of resources per elite to the point that some in elite status were effectively poor. Instead of money they got status and a title.

[1] https://en.wikipedia.org/wiki/Elite_overproduction

I see these claims pop up occasionally and I have found them both very overstated but potentially insightful.

With a perspective that goes farther back in time and a wider geography the sleep patterns promoted here as universal simply stated are not.

However, humans have always been flexible in their sleep patterns, and a lot of modern sleep pattern promotion is overly inflexible.

There are studies of hunter gatherers that lived close to the equator. Most of them did not nap most of the time, but in the summer they might nap during the day 20% of the time. They usually slept 6-7 hours in one stretch. [1]

But sleep was very flexible. In some cultures people would go to sleep at different times so there was always someone around the campfire. Mothers would nurse babies. Some cultures might for example once a month during a full moon stay up and party at night and then nap a lot the next day. And the sleep patterns mentioned in the submitted article show further flexibility.

Your sleep quality is likely a lot lower than a hunter gather due to modern light pollution, ability to use devices and have entertainment at night, and a lot of other factors. Unless you are monitoring your sleep with tracking devices there is a good chance you are probably getting a lot less (like an hour) than you might think. So committing to a 9 hour sleep window is still a good idea even if it might be natural in ideal settings to sleep for just 7 hours. Ultimately though you want to feel well rested with energy for your day rather than satisfying a belief system about sleep.

My approach with young children is to go to bed early so that if I get woken up I can deal with that and have time to go back to sleep. I might end up in the biphasic pattern of northern Europe sometimes with that. There was a time where I more intentionally tried to take a short nap during the day. But now I just take a nap if I feel tired.

References

  * [1] Hunter Gather Sleep study: https://www.cell.com/current-biology/fulltext/S0960-9822(15)01157-4
  * Why We Nap
  * The Old Way
  * Keep the River on Your Right

I think Spain is showing that majority renewables can work, if you are willing to work through the pain. Someone needs to lead the way and figure out how to make it work, so I am glad they are doing it. I do think this article and most information is focused on cheer-leading rather than telling it like it is.

This title of the article is misleading. Spain is one of the cheapest power generation markets, but In the article it states "Spanish households pay above the EU average". Then the reason is stated "Other system costs are rising. The flip side of getting energy cheap is paying more elsewhere to keep the system stable.".

Spain of course also had a blackout and the article states "every country in Europe needs to modernise how it handles voltage stability". I believe that's code for "its harder to manager power transmission grids with renewables". I have been told by a power engineer that read the report on the blackout that the authors are going out of their way not to explicitly blame renewables, but these things that caused the blackout are bigger issues now due to the switch to renewables.

But it makes sense for Spain to continue down this path and not pollute with coal or rely on other countries for gas imports.

I think 2) is a lot more complicated to the point statements like that are misleading.

Take a look Graph of energy consumption of China which is about double the US: https://ourworldindata.org/profile/energy/china

The energy consumption of the United States has flat lined: https://ourworldindata.org/profile/energy/united-states

One can argue that the US and Europe have maintained a low energy consumption by de-indusrializing and having China produce all the energy (largely with coal!) to manufacture their goods instead of manufacturing it themselves.

1) Is a lot more complicated as well. A simple ICE vs EV comparison ignores electric grid generation efficiency and transmission losses as well as the massive energy cost of manufacturing the battery.

These reports are inferring a lot from 1 year trends that are often changing only around 1%. Certainly it is great if new energy is coming mostly from cleaner sources, but the idea that we are actually getting rid of the non clean sources is something we should be skeptical of.

This graph shows all energy usage over time: https://ourworldindata.org/grapher/global-primary-energy

New energy sources have always been additive. We have never gotten rid of an energy source unless we exhausted the resource or it got prohibitively expensive (whale blubber having a population collapse). Coal is far more polluting then any other fuel source and globally we aren't reducing its usage. This graph is not updated for 2026, but I doubt the message will change much.

As we now undergo a worldwide population decline things might change. But at the same time we are also introducing energy intensive technologies: AI and robots, so there is no clear end in sight to increased energy consumption yet.

The difference is that Go has its own linker rather than using a system linker. Another article could explain the benefits of tighter integration and the drawbacks of this approach. Having its own toolchain I assume is part of what enables the easy cross compilation of Go.

Functional programming is immutable by default. TypeScript and many other typed languages don't really stop you from clobbering things, particularly with concurrency. Rust does. But immutability with GC is a lot easier to use than Rust if you don't need the performance of Rust.

The recent move away from phonics has been disastrous, and states that are using phonics now are seeing better results.

Some have called it the “Mississippi miracle” ...

A clear policy story is behind these improvements: imposing high standards while also giving schools the resources they needed to meet them. In 2013, Mississippi enacted a law requiring that third graders pass a literacy exam to be promoted to the next grade. It didn’t just issue a mandate, though; it began screening kids for reading deficiencies, training instructors in how to teach reading better (by, among other things, emphasizing phonics), and hiring literacy coaches to work in the lowest-performing schools. Louisiana’s improvements came about after a similar policy cocktail was administered, starting in 2021.

I would be interested to know more about the approach with literacy coaches. I donate to a charity that does 1 on 1 reading tutoring: https://readingpowerinc.org/

If we cannot as a society teach our children how to read, something is very wrong and we need to invest heavily in fixing it.

Meta Ray-Ban Display 10 months ago

Is there a way to just use this as a computer monitor? That’s what the Viture glasses are and it’s great to have a portable monitor that focuses at a longer distance.

Navidrome looks nice but it looks like it is Desktop only. I am using Plexamp as well. I tried some alternatives but couldn't get them to work reliably. People miss Plexamp as an option because they try the regular Plex app and not the simplified Plexamp.

I like Orion on iOS- it’s Safari without all the ads showing up. I can run the DarkReader extension at night just like Safari. Unfortunately it’s the most unstable software on my phone- at times regularly freezing up and then I switch back to Safari.

Texas is one of the best climates in the US for renewables but in locations with less sun and wind the math will be different. That math includes batteries for load shifting of which Texas is installing a lot.

As renewable generation increases past a certain level grid stability does require additional effort and that’s a lot more difficult to price in. In Texas their grid is isolated from the rest of the US. This may create a lower ceiling on renewables since they can’t send excess generation anywhere other than their own batteries .

This is one of the most disturbing articles I have seen related to reproducibility because it seems to imply that scientists don’t already know this.

One definition of legacy code I have seen is code without tests. I don't fully agree with that, but it is likely that your untested code is going to become legacy code quickly.

Everyone should be asking AI to write lots of tests- to me that's what AI is best at. Similarly you can ask it to make plans for changes and write documentation. Ensuring that high quality code is being created is where we really need to spend our effort, but its easier when AI can crank out tests quickly.