FlyingAvatar

I am curious as to your experience with this.

Over the years, I have administered a few dozen small to medium domains (depending on the domain 10s to 10,000s emails per month) and the only thing that has ever affected delivery is the reputation of the sending IP address of the mail server (and ensuring DKIM/SPF alignment in more recent years).

It's been quite sometime since I have heard the term "lamer", but I remember it being a ubiquitous insult in the BBS scene in the early to mid 90s.

Yes, the Agent mode is terrible. Have you tried using the Ask mode instead?

The Sonnet 3.7 'Agent' implementation in Cursor is terrible. It takes 10x longer to produce a worse response. I don't understand if this is a cost saving measure to reduce token count, but it is really bad.

Thankfully, switching to Sonnet 3.7 in 'Ask' mode has been essentially fine and is comparable to how Sonnet 3.5 performed for me.

Played again this week. I think the reason I missed the counters before is that they aren't adjacent to the play field and on a wide screen device, they are "way off in the distance", so to speak.

Cool, I will keep an eye on this.

You have two requirements that are at odds:

* Macbook is not an option

* Want to be able to mess around with some local LLMs.

Your choices for a Window laptop that can run a local LLM is either to get a large amount of system RAM and have it be abysmally slow, or to run a very tiny model on a discrete GPU which will (a) not be very good due to its small size and high quantization and (b) evaporate your battery life.

If you want to run local LLMs on a laptop and actually have them be useful, a Mac is currently the only real choice.

That said, with the money you save buying a Linux laptop instead, you can pay for a lot of tokens for whatever hosted LLM you want and it will be higher quality than what you could potentially run locally on a Mac.

I realize now this is a streak counter not a number of lives, which I think is even less intuitive.

Having lives, coins and streak counters outside the play area I think would be better.

Love the concept.

My $0.02 of feedback: The number of lives being inside the puck as a number is distracting and makes the game look less evolved than it actually is.

I would prefer a visual indicator on the side of the game play area.

I think it's an off by one error. When I got to level 2 of the perk, it ended when there was 1 block left.

Yeah, I think it would be a cooler visualization if the drives were in a line instead of a circle and new drives are added on to the right. The failed drives pile up on the bottom.

Still fun to watch as it is, though.

Indeed she was not apparently getting much better until she got the nerve block procedure a few months ago.

You have a choice of what public school your child goes to based on where you live. Many parents “shop” for their schools in this way.

Agree that it does make the product better, though if they put their RAM chips soldered on the board instead of in package, it would allow repair or upgrade without sacrificing the whole CPU unit.

GPUs do not put their RAM in the same package as the processor itself, and it does not sacrifice a significant amount of performance.

The M-series processors' primary benefit by a huge margin is its architecture that allows a higher memory bandwidth than say SODIMMs and that benefit is independent of the choice to put the RAM in the package.

I agree with your sentiment and wanted to point out that for the RAM it's even worse than the storage. The RAM is not soldered on the board, its included in the package with the CPU, which makes upgrading the RAM effectively impossible for these machines.

I don't have direct experience with long, straight cuts in my use case, but based on what I have seen in the forums the Maslow 4 is pretty good at them. It has an auto-calibration feature that the original did not have which makes it much less finicky to dial in than the original Maslow.

In the forums, I have seen people cutting cabinet parts with it, so I have to assume the straightness is acceptable, but it might be worth asking those who are using it for that purpose specifically.

I backed the original Maslow and also the Maslow 4, and recently a 4.1 upgrade. The creator is a really dedicated guy and the Maslow is a great CNC for those who cut sheets goods from wood (though folks have adapted it to other materials) and are OK with some tinkering.

There is a decent sized community that supports the software as well.

Yes, CloudFlare or similar.

I mean, I understand your sentiment on addiction but in the US, we have not been terribly successful in ending addiction despite most drugs being illegal for decades. In places like Portugal, legalization with an emphasis on treatment has not really been more successful either.

Having an attacker know the public IP of your box almost guarantees that they can take it down. Even if somehow you were able to ignore all the packets from the attacker, most capable DDoSers will be easily be able to saturate the bandwidth of pretty much a single VPS easily.

And if it's a cheap VPS, your provider is going to drop you in a blink so that their other customers are not affected.

If you want to be able to withstand a DDoS, you need a WAF who can absorb the amount of bandwidth that the attacker is capable of delivering, while also being capable of filtering enough of it that your service is not overwhelmed anyway.

Also, if you have ever hosted your service directly on the public IP of your VPS, you also probably need to change it. If an attacker sees your service is suddenly behind CloudFlare, they can search sites that keep a history of domain's public IPs and will attack all of them to see if they can skirt around your WAF.

Ideally, you need to keep the IPs of your services completely private and then firewall them such that only your WAF provider can forward traffic to them.

This is exactly how I handle my mail as well.

I would argue the parenting is of much greater value than the birth itself. In general, the better the parenting, the greater the net value of the child to society.

> We conclude that the US government should be willing to pay up to $290,000, overall, to each parent who gives birth.

This is a weirdly short-sighted suggestion after all that analysis. Getting paid to have children can easily cause the value metric per child to plummet; not to mention the probable suffering caused to the children whose parents might try to abuse it.

If you want to increase births, things like subsidizing daycare and after-care and taking steps to ensure the quality of care are helpful. Making the cost of having children lower indirectly is much less likely to be abused.

Additionally, if you want to increase the value of children to society, making teaching a more appealing profession (i.e. paying more) and taking steps to ensure the quality of teachers are helpful.

I have been using Zed for the last several months, and their new Zed AI addition is quite good, which I think is only available with their Preview (i.e. beta channel) release this far. It also comes with free in-app access to Claude 3.5 Sonnet (at least for now).

It really depends on what you need to do with the data, but in most cases Python could do this pretty easily with csv.reader (with a \t delimiter for TSV) or xml.etree.ElementTree.iterparse (for XML) in streaming fashion such that you're not loading the whole file at once.

Haven't USB hubs done this since USB 1.1?

How many jobs have you actually applied for? Do you not get any response at all? If you've applied for more than 20 jobs or so and have had no response, I would ask folks to review your resume here.

Serious questions: If there were so many of them, how did the justice system not figure out this was flawed much sooner? How did the legal representation for these subpostmasters fail to be able to prove their innocence?

This happened over 20 years ago, but I was helping a co-worker debug an issue they were having with a Windows application written in Delphi. This was before Google was a thing, and waaay before Stack Overflow, so getting help to solve these kind of problems was a bit more involved.

As far as the issue, if they ran the offending code in the debugger, it worked flawlessly. But it would fail every time in the production build. Usually, this would point to some kind of race condition, but the code section was innocuous. It was essentially running the Delphi equivalent of strpos on a local variable.

I was comparing the build flags between the debug version and the release version and one thing that caught my eye was the optimization flags for the compiler. Lo and behold if you brought the optimization level down two notches the bug went away.

I don't think I ended up getting into the disassembly to submit a bug report, as the optimizer was almost certainly doing something it shouldn't, but at least we found the source of the issue.

Since we didn't want to actually disable optimizations on our release build, the "permanent fix" was to re-write our own strpos-equivalent in such a way that compiler optimizations didn't break it.

I switched from Chalice to Zappa several years ago and now manage several production projects with it.

I still have one of those Gorilla/Banana printers. Not that I actually use it.