CWuestefeld

In the bigger picture, people were starting to catch on to the fact that Google's search product was becoming increasingly useless.

Around here we'd come to that conclusion at least a couple of years ago, due to abusive SEO and so forth. And that understanding was becoming even more widespread.

I don't know if Google's got the later parts of the game figured out yet, but I have to think that they'd realized that Search was dying. While there's still some value there, may as well use it as a hook to pull people into what they expect to be the next era.

I'm not at all an industry pundit. But I suspect there's a reason we're not seeing leading models from Google recently.

Judging from my own frustrating attempts to use Gemini for vibe-coding, it seems like Google is badly over-sold (i.e., under-provisioned).

From all those promos giving away their pro-level subscription with phones; spinning up a mid-level subscription to undercut other providers and (probably most significantly) putting AI queries into ever search response because their flagship search product had become useless; they're promising a lot more processing to customers than they can reliably deliver.

The recent iterations seem to be intended not to push the capabilities forward, but to deliver capabilities at the current level while consuming less resources. That will allow them to maintain their trajectory until (I'm expecting) they get the huge infusion of extra compute resources from Space X later this year.

If I'm right, then I expect we should see Google start pushing forward again (rather than more of this lateral stuff) by the end of the year.

I'm pretty sure you're misunderstanding the definition of "targeted". It has nothing to do with the total capacity of the system, it's about the ability differentiate a specific target from everything else.

The design of these systems is the opposite of that: they're intended to intercept everything within their reach. The fact that with only four devices, their reach isn't (yet) so great has no bearing.

My Instinct 3 Solar shows 34 days remaining in the battery. I charged it sometime last week, so we're talking about upper 30s at least.

That's with running activity tracking including GPS for about an hour a day, and continuous heart rate monitoring.

At least on my device (and my body) the spo2 sensor is very finicky, requiring absolution stillness, and even at that I don't think it's accurate. So I don't use that.

I also discovered recently that simply using a custom watchface - even one built with an eye toward efficiency - lowers my battery life to about 2 weeks. The native faces enjoy privileged operations and are thus highly efficient. But custom-coded stuff apparently runs in a sandbox that has the effect of costing much more power. That's unfortunate, because my custom face used custom high-legibility fonts (derived from what's on interstate signage), making it noticeably easier on my aging eyes. The built-in fonts aren't great for this.

Garmin watches that have mapping support have an introduction price of >600 Euro. Even at that price point, zooming or panning maps is excruciatingly slow

I just got a Garmin Instinct 3 Solar. It does mapping, and cost me about $300 US.

You're right that it's slow due to a wimpy processor. But the processor isn't because they're too lazy to innovate, but because they have something sipping tiny amounts of power so that I can get a battery life of several weeks.

I want to support the "what about debouncing" argument mentioned elsewhere; the author shouldn't just ignore this.

But I also hate the "you had one job" meme and want to argue against its mindless usage. Most of the time, when people do the "you had one job" thing, it's false. And that's true most of the time in the case of buttons, too. In a typical user interface, a given button has some combination of these jobs:

* Communicate what action will occur should the button be pushed.

* (Sometimes) communicate the current status of some aspect of the system (e.g., often a button is used to enable/disable a mode, and the button itself visually conveys what the current mode is).

* Execute the intended action upon clicking.

* (Sometimes) communicate that the command has been received and is being executed (e.g., in the OP, the button might disable itself while animating the rotation in order to avoid the confusion the OP complains of).

I haven't played this, but just reading the description...

Learn to hack from TRASH WORLD NEWS, the underground computer magazine.

It seems like a missed opportunity not to name-drop 2600. But I guess they wouldn't be allowed to do that anyway.

There are no cars - or roads - where the lights can be seen.

I'm aware that there's widespread support for the idea that there's some odd atmospheric effect with refraction or something, that displaces the apparent location of real headlights. But even granting that, it doesn't change that it's a pretty weird thing.

The scarcity comes from there only being a finite number of the devices.

But I was really speaking more broadly, about healthcare in general. There are only so many doctors, so many MRIs, so many heart surgeons, so many this or that. Giving everyone a bunch of extra money so they could pay the current prices doesn't do anything to address the scarcity, so everyone flush with those subsidies will just bid up the prices - which is exactly what happened.

We went there a few years ago, while we were in Fort Davis, to see the Marfa Lights. We did see the lights - it's a real thing! But at the same time, we also saw a Space X launch, including seeing the flash of the 1st stage separation. It was an amazingly cool experience.

While waiting for sunset, we also took a drive out of town to the southwest toward Mexico. It started as a nice paved road, but after an hour or so the pavement disappeared. A bit later, we started meandering down the side of a mountain, on loose shale, only one lane wide and nowhere to turn around. The views were amazing, but it seemed that was a clear sign that it was time to turn back.

Scarcity demands some means of rationing out the product. Setting higher prices is one means of doing this, so only those with some means of paying can get it. Another approach is via wait times, where only those who can wait and afford the time penalty can get it. There are other variations, but there's no such thing as a free lunch.

One of my pet peeves is what seems to be an overwhelming desire in writers to always put an adjective in front of every noun. You can never just let it be a "notebook", it has to be some kind of notebook.

It's even worse in product naming and advertising. Nothing can be just "vanilla", you have to even put an adjective in front of your adjectives, like "Mexican vanilla".

EDIT: s/verb/noun/

I was just making the change from LP to BW yesterday, completely by coincidence. My first reaction is that the out-of-box experience is poor.

The first step was easy. The account creation and import of legacy data all went pretty well. But after that it wasn't so pretty.

The first hurdle was trying to understand their model for sharing data (so my wife and I can share important credentials). The model that LastPass uses is pretty intuitive to me: it's just a matter of sharing a folder, so relatively transparent. But Bitwarden has a whole separate concept of "organization", and the items being managed don't go in "folders" here, but in "collections". So there are two separate, and subtly different, models in play, and this is confusing. The good news is that the client aggregates the data so when you're using it day-to-day to fill login forms, you don't have to worry about the differences.

Once I'd gotten the data in place, I had to get the clients set up on the various platforms (browser extensions; desktop native, which is actually required for the browser extension's security to work right; phone). The OoB settings were entirely paranoid, and had me re-entering the complex master password over and over, really annoying me. Figuring out how to get to a reasonable balance required figuring out some settings whose labels are misleading. For example, "Unlock with PIN" sounded to me like it was going to add an extra layer of security, but it turns out that it really means "allow unlock using PIN in lieu of master password".

Also, note that while most of the settings default to paranoia-level (like the "require master password every time I inhale", that I mentioned above), you will probably want to change the default crypto cypher. It defaults to PBKDF2, but a better modern approach is the other choice, Argon2id.

...which also reminds me that there's a distinct lack of parity between client platforms. Although you need the desktop native app to manage browser extension security, there's a bunch it can't do. For example, after importing my legacy data, I needed to select all the contents of my LP shared folders and move them to the BW organization collection, but the native app (which seems to be an Electron app, btw) doesn't have a multi-select feature; you need to do that in the online web app.

If you like this, then try playing Red Dead Redemption 2. While just a tiny part of the game, you improve your relationship with the rest of the gang by doing chores like splitting wood, and also carrying hay to feed the horses.

I mostly agree with your criticism of my post. I was being generous trying to avoid being inflammatory here, since I know there are readers that strongly support socialist ideas (in the strict sense, not just the "safety net" sense). It was certainly Marx that pushed it so hard.

But researching this a bit, I find that it still predates Marx. I find:

Sir William Petty, 1662: "If a man can bring to London an ounce of Silver out of the Earth in Peru, in the same time that he can produce a bushel of Corn, then one is the natural price of the other."

More important, it seems that David Ricardo (a big name in economic history), in 1817 latched onto what Smith had written and states it quite definitively.

an honest and accurate correlation between display of effort and value

Hmmm. Your choice of words here has just sparked a realization for me.

Before you said this, I was completely on board with the original post. But in juxtaposing effort with value, it illustrates that we're basing the idea on the Labor Theory of Value. That idea seems intuitive, and Adam Smith wrote about it 250 years ago. But it turns out that LTV is very wrong. Economists showed that effort does NOT impart value.

This has been my policy for a couple of decades. When somebody posts just a bare link (especially if it's to a video), I refuse to click. If it's not worth your time to introduce why something is relevant, then it's not worth my time to go figure that out.

I acknowledge that the airline captain has some responsibility for our security. But part of this responsibility is being a steward for our overall well-being. And in this case, the "security" aspect is so vastly overwhelmed by the damage it did to passengers in other ways, that it was obviously a bad call on the captain's part.

It really does break both ways. Over-reacting to perceived threats has a cost too.

Warning - semi-political (but hopefully non-partisan) political content ahead: This is the same thing the FDA does with drug approvals. They are overwhelmingly biased toward preventing bad drugs that they prevent access to a lot of things that could help. Studies show that the FDA's difference between up-side and down-side risk costs a lot of lives on net. For example, the FDA delayed the approval of beta-blockers (used to prevent second heart attacks) for several years after they were widely available and saving lives in Europe. Analysts estimate that this delay alone cost tens of thousands of American lives.

Sometimes, accepting a risk provides the greatest net benefit.

I think godelski is being far too permissive in his odds. As he says, we need to examine how (un)likely is it that somebody is trying to execute this terrorist action, including being competent enough to create a workable bomb, to sneak it through security, and so forth. That's all his numbers show.

But we've also got to factor in

A) How likely is it that this bomb is going to have some bluetooth component? It seems like needless complexity, so we should weigh strongly against this. Further, it's less likely that our hypothetical terrorist needs to have expertise in this domain as well.

B) How likely is it that he would clearly paint the word "BOMB" on the side of his device (figuratively, of course, since this is digital)? That's amazing levels of stupidity. And then intersect that with the claim that he's competent in all the other things (bomb making, sneaking through security, making a bluetooth trigger for his bomb) but is so incredibly stupid that he'd label it a bomb.

Factoring all of this in, godelski is being far too generous in assessing this with odds similar to finding the winning Powerball ticket outside the front door while simultaneously being hit by lightning.

I acknowledge that the airline captain has some responsibility for our security. But part of this responsibility is being a steward for our overall well-being. And in this case, the "security" aspect is so vastly overwhelmed by the damage it did to passengers in other ways, that it was obviously a bad call on the captain's part.

Essentially, put in the effort and do the liquid bowel prep.

It's not just about effort. I must do the liquid prep due to my Crohn's disease. And while I am able to get the liquid down (as you note, it helps to make it as cold as possible; also, suck on an ice cube before drinking to numb the taste buds), I can't keep it down. Within an hour it has me evacuating from both ends.

For my last test, I barely slept at all the night before on account of the vomiting, and even once I got to the hospital I was lying on the wonderfully cold tile of the floor between rolling over to vomit in a trash can.

They know it affects me badly, but still assess that it's necessary due to my risk factors. And because I'm losing much of the drug due to the vomiting, the prep is poor, so I have to start fasting a day early to ensure that I get sufficiently cleaned out. It's torture all around.

I generally agree with your point about it being simply emergent behavior.

But on the other hand, the timing (having seen over the past week or so several articles about the most disastrous IPCC model now having become implausible) makes me wonder if some individual actors are thinking they need something to shore up their disaster prophesying.

Trademarks are a fundamentally different kind of IP.

With copyright and patent, the creator of the work is being protected. But with trademark law, it's not about protecting the content of the IP as such. It's about protecting the consumer from being misled into thinking they're getting the real thing.

And given the guitar market at large, with about ten thousand different guitars in the general shape of a Strat, it's pretty much universally known that the name on the headstock is what you have to look at to differentiate. So long as that name isn't misleading, I have a hard time imagining how they could make a case of it.

I mean, if the headstock says "Fernando Stratoblaster" or something, then MAYBE it's a little confusing. But my guitar, a Kramer Focus 6000 looked very nearly identical to a Strat (the edges are less beveled, the headstock is pointier, but at a quick glance...), but it quite clearly says that it's NOT a strat. Nobody's going to be fooled despite the striking similarity in shape.

I assume that the repository of books was used as training data, but not by way of the annas-archive domain. Instead, it would make a lot more sense for them to download the whole pile via bittorrent, which has nothing at all to do with the domain. In other words, the legal solution here wouldn't have prevented the problem.

This was referenced at the bottom of the linked article.

And yeah, I'm a big fan, too. I still have the CDs for it, and it still runs in Windows 11!

So they brewed up a bunch of ugly C macrology that enabled C programmers (or Visual Studio wizards) to define COM interfaces in header and implementation files that just happened to lay out memory in the exact same way as vtables of C++ pure virtual classes.

While C++ programs would use other ugly macros to declare actual honest-for-god C++ classes to implement COM interfaces.

And Visual Basic programmers would ... do whatever it was that Visual Basic programmers did.

We were being a well-behaved MS shop at this point. The new generation of our website was designed with a distributed component architecture being called from ASP (Active Server Pages) using DCOM. We implemented it in C++ using ATL (iirc) to facilitate the COM interfaces. The thing was, even with that help, our actual business logic was absolutely overwhelmed by all the annoying casting back-and-forth between those IUnknown values and C++ native types. It was really annoying.

This was around the timeframe of VB 5, I think, and we discovered that we could write the same logic in VB without all the annoyance. Putting aside our "serious developer" C++ elitism, we were actually productive in doing it all in VB instead of C++, and that generation of the system lasted for a bunch of years. We eventually replaced that with an ASP.Net design in C#, which was a whole lot more manageable.

Speaking of single datum...

It's $4, the reviews are good, just buy it.

Yeah, $4 isn't relevant. What is relevant is the time investment, which your response ignores.

I recall an internal (never released) project at IBM in the late 1980s. It was a tool for creating client-server GUI apps, programming them with the REXX language. You may remember that client-server was all the rage at the time, and REXX was IBM's favorite scripting language. IIRC, the internal name of the project was "Red October", but I can't find any reference to this online.

The tools lacked the visual GUI builder of VB, but really, that's just a detail. The rest of the framework was really quite powerful, and a GUI builder could have been added. But in true IBM fashion, they had no idea how to market something that wasn't mainframe targeted, and they killed the project. There was a fair amount of acrimony on internal forums about this at the time.

I think you mean "Envelop" (note the lack of trailing "e"). That was a neat tool, and community, back in the day. Too bad it didn't get anywhere.

I've seen that, but always assumed it would be a more accessible account, and not tickle the technical developer itch.

I recently read Jordan Mechner's The Making of Prince of Persia, and while it was interesting, it wasn't what I'd hoped for. It was a biography covering his life during the creation of the game, and very little about the internals and technology of it. That's fine, but for my tastes, the book that it really was, was less captivating than it could have been.

My default is to expect the same from the FoxTales book.

VB was never included in Windows (although it did include the runtime files).

Windows did include other BASIC-related stuff, like GW-BASIC, QBasic, and VBScript.