Agreed. The idiotic law not allowing college debt to be cleared by bankruptcy is the primary reason why college has gotten so expensive.

Then treat college debt like any other loan instead of subsidies backstopped with government bailouts.

> The inevitable and unfortunate result of "handing out 'free' stuff" however will always be widespread fraud.

The optimal amount of fraud is not "zero".

The goal is to enforce just enough to keep the fraud below a threshold while not making it too onerous for the people who need and receive the benefit.

And, if you really want to go after fraud, you are far better going after the people who commit it at an organized institutional level: see GOP leadership in Florida and Medicare.

> There should be volunteer groups at local libraries running these services for their local communities.

The problem with bespoke anything in computers is always the support.

No one wants to be on the hook for customer support. I absolutely agree with them.

There are a ton of "services" that exist solely to enable people to cut a check and say "Customer support is over there. Go talk to them and leave me alone."

> I was fighting deadlocks in some Java code this week

Why? You have java.util.concurrent; you should never see a deadlock. You might see a performance degradation or maybe even livelock, but that's very, very, very rare.

What abjectly idiotic thing is in your Java codebase such that you have deadlocks?

> The models are nondeterministic, and therefore it's pretty normal for different runs to give different results.

And how is that an excuse?

I don't care about how good a model could be. I care about how good a model was on my run.

Consequently, my opinion on a model is going to be based around its worst performance, not its best.

As such, this qualifies as strong evidence that Opus 4.6 has gotten worse.

>> The models are nondeterministic, and therefore it's pretty normal for different runs to give different results.

> And how is that an excuse? […] this qualifies as strong evidence…

This qualifies as nothing due to how random processes work, that’s what the gp is saying. The numbers are not reliable if it’s just one run.

If this is counter-intuitive, a refresher on basic statistics and probability theory may be in order.

> If this is counter-intuitive, a refresher on basic statistics and probability theory may be in order.

I'm not running "statistics". I'm running an individual run. I care about the individual quality of my run and not the general quality of the "aggregate".

The problem here is that the difference may not be immediately observable. Sure, if it doesn't give a correct answer, that's quickly catchable. If it costs me 10x the time, that's not immediately catchable but no less problematic.

No, what they're saying is the previous run could have just been lucky and not representative!

> I really wish AMD and Intel boards get replaced by competent people.

Intel? Agreed. But AMD is making money hand over fist with enterprise AI stuff.

Right now, any effort that AMD or NVIDIA expend on the consumer sector is a waste of money that they could be spending making 10x more at the enterprise level on AI.

How is the cleanup with DrinkMate? Washing everything was my primary problem with SodaStream.

Not hard? I mean you need a dedicated bottle brush, but the bottle itself is the only thing I recall having to clean. The DrinkMate advertises itself as being able to carbonate other things (juice, flat soda, etc.), but I never carbonate anything other than water in the bottle, then add my syrups only after decanting. It prevents the bottle from getting sticky.

> The much bigger danger for home distilling is fire, as you have open flames and combustible vapor.

This would suggest that using induction heating would be significantly safer and have the possibility of precise temperature control. Is there any reason why home distilling does/does not do this?

Electric heating does reduce the risk of fire, yes, and some of us do it. (It’s also just a lot easier than a turkey fryer.) I rigged a water heater element up for this purpose.

(Technically there actually isn’t temperature control in distilling, the temperature is just the boiling point of the mixture, which changes over time as the mixture changes from distillation, but you do control the heat input which effects the speed at which you distill. Tangential, but counterintuitive.)

The reason most don’t is just cost/practicality. You really need to have a fair bit of liquid to get good results. Like tenish gallons (~40L). You probably can’t fit a still that big on your stovetop (and you really want to do this outside anyway) and you’d need a 240v connection to provide enough heat. Your standard American wall outlet doesn’t provide enough juice.

But the standard 240v 50a you charge an EV with or, in my case, plug in your RV does. People run drier cords out a window too.

> Like tenish gallons (~40L).

Ah, that would do it. I was thinking this was like beer homebrewing and would be around a gallon.

Thanks for the info.

Yeah, the thing is as you distill you’re saving it bit by bit as you go along. You toss out the very first stuff (called foreshots) because it contains a number of chemicals with lower boiling points you don’t want (methanol, acetone, etc.) in higher concentration.

Then you get the heads, hearts, and tails and blend them together according to taste. You just wouldn’t get much separation if you distilled a small amount unless you were collecting in really tiny quantities.

So it just becomes harder to do a good job with a small amount.

> Will RISC-V end up with the same (or even worse) platform fragmentation as ARM?

Sadly, yes. RISC-V vendors are repeating literally every single mistake that the ARM ecosystem made and then making even dumber ones.

Please elaborate.

> In any build system, I think the distinction between “cross” and “non-cross” compilation is an anti-pattern.

This is one of the huge wins of Zig. Any Zig host compiler can produce output for any supported target. Cross compiling becomes straightforward.