diggernet

Oh, this is great work! The flOw game finally works!

Archive.org [1] is still using an older version of Ruffle where it is broken, but I found another site [2] using the current Ruffle and it even works on mobile.

[1]https://archive.org/details/flash_flow

[2]http://www.xgenstudios.com/game.php?keyword=flow

Obligatory mention of the waterproof Galaxy S5, with its quickly-swappable battery by popping the back off.

If you are willing to trade heat and pry tools for thinness, that's a legitimate opinion, but please don't blame the trade-off on waterproofing.

Perhaps it should be a part of management ethics courses.

My 10 year old Thinkpad with S3 wakes up in the time it takes to open the lid, with almost as much battery as it had when I put it to sleep days ago.

My shiny new work laptop without S3 wakes up in the same time, and then says "Battery critically low, shutting down."

This is not a convenience, it's insubordination. I told it to sleep, and it didn't.

But you know what's really insane? When the new work laptop is just sitting there with the screen locked and I walk up and hit a key, it takes longer to just display the password box than the 10 year old Thinkpad takes to wake up from S3 and display the password box. (Ok, it's an unfair comparison, Windows vs Linux, but still...)

Not to mention the opsec failure of the starlink antenna continuously broadcasting the ship's location during its deployment.

I have an HMD-made Nokia XR20. Rugged, waterproof, headphone jack, dual SIM or SIM + microsd, unlocked, Android One (little or no vendor or carrier bloatware). Really great phone.

It's not a problem with a camera based system. It's a problem with an overly complex AI+camera based system. If the car has to recognize what the camera sees before it will react, it's going to hit things. But you can have a camera based system with good binocular vision to build a depth map, no AI stuff. It doesn't matter if it's something seen before. If something is there, you don't need to know what it is, just don't hit it. The AI stuff is helpful for a lot of self driving decision making, but detecting objects for collision avoidance needs to be simple and reliable.

The site was built to survive a nearby nuclear attack.

All of those can be done today, because none of them benefit from the TV being transparent.

Yikes. Stupid phone thinks it can spell better than me. Fixed. Thanks.

The problem with .us here is that private entities can register domains there.

I think this change is great, because when you are looking for official information for a city or county and all you find is a .com, there's no way to be sure it's legit. Using .gov signals it's official, but .us does not.

Yeah, that was the line where I started to lose interest in the post. I even stopped and reread that bit several times, trying to find any positive interpretation to balance the huge negatives.

Doesn't that simply turn the "other information channels" into another advertising channel? If you are putting out codes to bring in customers, and customers arrive with the code, um, mission accomplished?

Beyond Compare

https://forum.scootersoftware.com/forum/beyond-compare-4-dis...

Oh, that's a good start. Thanks! Though from what I've heard, Google won't allow replacing the RCS app like you can with SMS/MMS. Which would significantly limit its usefulness, unless that changes.

When can we have a non-Google RCS app on Android?

One of the ways that tickets are a deterrent is if you get enough of them they revoke your driver's license. Seems to me that it would be a pretty big deal for the corporation if its "driver" could no longer legally drive any vehicle!

Archive of the original page:

https://web.archive.org/web/20230522141639/https://www.rickg...

The message about stolen artwork is most unfortunate.

Please pardon my failure to remember the correct term. But fundamentally, it constructs a 3d model based on what it thinks each pixel represents. And sometimes gets it wrong, leading to it having a pattern of driving into stationary objects.

The problem is that Tesla uses an ML classifier to identify objects it sees. Once it has identified them, it can estimate how big it is, and thus how far away it is. But if the classifier doesn't identify the object, it might as well not exist.

Contrast this with Subaru, who uses binocular view through two widely spaced cameras to build a depth map of the view in front of it. It doesn't know or care what is there, only that something is there it shouldn't hit. And so the emergency breaking will trigger for things like trains and stationary emergency vehicles that Tesla seems oblivious to.

Sometimes simpler really is better.

It's only not a security problem if the file never leaves that single user system. Which was definitely not a safe assumption, even then.

Is there an integration for it? And can you set it up completely offline, without needing to create an account or install an app?

Because you are afraid of being dropped out there into a months-long grind to find anything. I'm saying that if tech companies hired like normal, sane companies, that would not be a fear. Sure, they might not keep you through probation (and YOU have a lot of influence over that). But if not, you go start at another company. That's the benefit I'm talking about with not having stupid long, arduous interview processes: Not having stupid long, arduous interview processes.

The "standard" probation period is 90 days. Outside of the modern silicon valley bubble, lots of companies hire with a probation period. Inside the bubble, the equivalent is contract-to-hire. But I really wish companies would just go back to the old days of probation.

My current job was contract-to-hire. Officially an 18 month period, but they converted me to employee after 3 months.

How it helps the candidate is exactly the same as how it helps the company: The interview process was a half hour phone screen, a 1-2 hour onsite with two people from the team, and a half hour followup call with the hiring manager who hadn't been able to join the onsite.

You know those week-long all-day interview processes many of us hate? Not necessary at all if the company is willing to give people a try and just move on if it doesn't work out.

You know how companies complain that they can't find good hires because their ridiculous interview process filters out the good ones and keeps the great ones from even applying? Having a sane short interview with probationary hire costs way less productive time and gets people onboard fast. And if someone doesn't work out, you can have another candidate take their place next week. It doesn't need to take months.

Don't worry, Google is working on that.

The demise of email forwarding is getting closer

https://news.ycombinator.com/item?id=40045069

This is really cool, but usage looks very complex. I keep hoping to find an easy-for-the-casual-user tool that takes photos and videos and produces textured models. Kind of like Canoma was, back in the day.

Those are fun thoughts.

I've imagined a world where you can create a long, complex prompt that results in a reproducible response of code that does exactly what you want, and does it correctly. In that world, the prompt could be your versioned source and the build process could start from that. Implementing changes could be done by tweaking the prompt. Of course you'd still want to archive the "intermediate" source code, if only to protect yourself from an LLM outage.

In that world, you actually would be getting new code from the source prompt regularly, and yes, it hopefully would improve over time.

In our world however (at least for me), usage tends to be more conversational, so there really isn't even "a prompt" to save. Not to mention that I never use the output as-is.

Yes, the prompt is copyrighted, just like your post here is copyrighted. But I don't think there's any point in versioning the prompts. The responses are just too nondeterministic. If you feed it the exact same prompt again, you can get something quite different back. On the other hand, if giving the same prompt got you the same response (like a reproducible build), then absolutely yes, version those prompts.

The top rate was that high, but nobody was really taxed at 94% because there were a lot more exceptions. When they drastically reduced the top rate, they also eliminated a great many exceptions. Overall, the change was actually revenue-neutral. Which shows that you cannot compare that 94% number to the current rates and say that people used to pay more taxes. "Lowering the rate" was just simplifying things so that the printed rate more closely matched the reality.