pcollins123

Being a bad manager. That’s a very personal attack.

4 days ago Grok 4.5 and GPT5.6 were released. In our study, they beat all Anthropic models for finding vulnerabilities in Pull Requests.

We ran 10 models over the same 10 pull requests, each carrying one planted access-control bug (IDOR, missing auth, broken authorization), five times per model, and scored every finding against the actual code.

The result is a clear cost/quality frontier. Open AI’s GPT-5.6 Sol dominates. It found all vulnerabilities at a very affordable $0.70 per PR. For comparison Fable only found 88% and cost $3.61 per PR. No Anthropic model reaches the frontier at all.

This is NOT a harness quirk, we tested and found similar results in the Claude Code and Pydantic harnesses.

Details of our methodology and synthetic repos in the publication.

100% agree! I built Paisley (because it is the opposite of plaid), to host your personal finances locally and is 100% open source. Paisley pulls data from your financial institutions by scraping balances and importing CSV exports, storing everything locally in a simple SQLite database.

https://github.com/patrickcollins12/paisley

You could summarize that data using some form of machine learning. A good new skill to develop. Then you don’t need to share the exact details, just a count per category. E.g. personal incident (32), late take off duty to X (23), passenger medical incident (15). Hopefully in aggregate form that data is less of a privacy issue and less of a commercial risk for your company.

Am I the only one who found this article a rambling mess? The first page makes a point, then it diverts off into a ridiculous rant.

This project describes how to build a quiet $20 PID-controlled thermostat controlled fan for cooling your media console or gaming cupboard.

The software is ESPHome and Home Assistant. The hardware is an ESP32 with a regular 12v 120mm Computer Fan (PWM) and a simple Temperature Sensor (DHT11).

Just bloody shootin for bronze… lookin for a Steven Bradbury.

With a $10 18650 rechargeable battery you could probably get a couple months on a single charge. Esphome can deep sleep between readouts: https://esphome.io/components/deep_sleep.html Deep sleep pulls the power draw down to 50uA.

Of course if you also wire a solar charger next to the battery… maybe it will never run out.

Seems like it worked for them.

Thanks Drew Houston, CEO of Dropbox!

Check your video card chipset and call support :P

Except when they're actively expounding upon and peddling bad news to ensure the stock price drops at the right time. That is what Musk is most frustrated by.

In my mind, all negative news about Tesla is now called into doubt as being news peddled by shorts. This is the challenge of declining journalism standards: all news now has someone with a vested interest behind it.

Even that was close, and the Republican candidate was _accused of molesting children_.

There, fixed that for you.

Look, I'm very progressive, but we need to firmly stand behind "innocent until proven guilty" especially when it is so easy to create a slander campaign against a political opponent. And, yes, even the democrats are capable of that.

For every service there are very relevant factors about what makes the product good or bad. If you don't separate these 2 or 3 factors, then over time everything becomes a score of 3.6.

In the case of Amazon, the relevant options are:

   1. Likert scale of quality:
       a junk, just don't buy it
       b cheap and works good enough for occassional use
       c higher quality: willing to spend more and you'll get a much better outcome.
       d overpriced

   2. bad shipping, bad vendor, poor customer service

In the case of rotten tomatoes it is, again a different set of parameters.

Google is releasing 20M bacteria-infected mosquitoes in Fresno... wearing small cameras and a projector that can display text advertisements

I'm not sure they will ever implement it actually. Nginx does not support HTTP/2 server push as of March 2017 (v1.11).

Owen Garrett at nginx has summarized the reasons as follows:

- it is a non-essential and optional part of http/2

- if the client already has the resource cached, then by pushing it to them you might be unnecessarily wasting bandwidth.

- server push spec might change in the future.

- Link headers as hints is useful, but usage has been low from web developers.

- server push has been available as part of SPDY and was not utilized by many web developers.

Read his original comments here[1]. This table[2] accurately describes the pro's and con's of server hints vs server push.

Personally this disappoints me because I think this would be a valuable feature for web developers willing to invest the time to optimize page speeds. I also know that some large CDN's like Cloudflare have implemented their own version within nginx to optimize page downloads.

[1] https://www.nginx.com/blog/http2-r7/

[2] https://www.chromium.org/spdy/link-headers-and-server-hint

This is really interesting thanks.

Doesn't this lead to the most highly used static assets being pushed for every subsequent request? Isn't that wasteful?

I use varnish cache extensively and am considering writing a probabilistic server push module. It could keep track of subsequent requests and push the next set of assets. This is more of a server module approach rather than client side.

Any thoughts on that?

Best ways to get into hobby game dev these days is PixiJS for HTML and Swift or Cocos2D for phone dev. Unity for 3D, but it is pretty hardcore. Get a good "learning 2d game programming book".

I learned SpriteKit and Swift for fun and started making games with my kids. So rewarding and a nice break away from real world programming. But not something I would want to do for a living.

Fire Particle Simulator with JS and WebGL https://ethanhjennings.github.io/webgl-fire-particles/