Onno (VK6FLAB)

Are you turning it the right way?

I’m asking because I’ve seen circumstances where everyone was convinced that it needed to go one way, but many expensive hours later were proven to be wrong.

Also, is there something jammed in the pipe, like say roots or a snake?

It’s not real.

At one point, before we virtualised everything, I had a custom desk built in an L-shape. Instead of a desk and a return, I had the refurbishment team put together a desk with two desks instead. It gave me two sets of drawers, two computer cubby holes and the gap was too small for the horrible keyboard adjustable shelf that kept hitting your knees, so they replaced it with a fixed surface instead.

People laughed.

Colleagues sniggered.

Then they wanted one too.

Now I have a mobile lectern with an iMac clamped to it. Height adjustable, wheels, enough space for keyboard, trackpad and USB hub. I move around my office as the mood or light takes me.

So, when you use 40 or so programming languages, your employer needs to supply a mansion…

I’m okay with that.

Now, where is the boss?

I’m going to answer your points below. Not because I want to tell you to move to Linux, but because the information you state is incorrect. Linux is not for everybody. It works for millions of people and it works for me, but that doesn’t mean it will be what you’re looking for.

In order:

1. There are no .exe files. Neither are there any on MacOS, iOS, Android, or anything else that isn’t Windows/DOS. To start software requires that it’s on the search path in exactly the same way that Windows requires. You can see what that is with the command: echo $PATH. Most Linux distributions have a graphical user interface which features icons and menus, but if you don’t want that, you don’t need to install it.

2. You absolutely can, but it doesn’t work the same way as Windows, because it’s not Windows. You can for example login to Linux because the login manager started at system startup. You see a desktop after logging in because there’s a startup system for your account. The printer works because the software driving the print queue is started.

3. Wine is a tool. It’s not a replacement for Windows. It’s not intended to be. It’s intended to help users and developers make Windows software work better on Linux.

4. LibreOffice is one of many office suites. I have been using it as my productivity software for 25 years in my company and I’m not at all disappointed to have escaped the Microsoft Clippy, Ribbons, Office365 abominations.

5. I have used Libre Calc for most of my numerical analysis processes. I used real tools like R and gnuplot when I was analyzing terabytes of data.

6. The terminal is a tool. I use it daily. At any time there’s a dozen of them open. Not everyone needs a terminal, but there are plenty of things that you can only do in a terminal. A random example, list all the files in your account, group them by extension, then add up how much space each extension takes. In case you’re wondering:

find ~ -type f | egrep -o "\.[a-zA-Z0-9]+$" | sort -u | LC_ALL=C xargs -I '%' find . -type f -name "*%" -exec du -ch {} + -exec echo % \; | egrep "^\.[a-zA-Z0-9]+$|total$" | uniq | paste - -

Source: https://unix.stackexchange.com/a/457241

Linux is not Windows. It never was and it never will be, neither is any other operating system. The community around Linux is helpful, the ecosystem is vibrant and it’s free. If you want to pay for support, you can. If you don’t, there’s plenty of opportunity to do your own thing.

If you want it to be like Windows, you’re going to be very disappointed.

Also, the word you’re looking for is: “headless”, as in, “headless install”

The traditional way is to use a serial console from another device.

What is it specifically about Linux that doesn’t work for you?

I’m asking because I’ve been using it for almost a quarter of a century as my main desktop.

Back propagation happens during the creation of the model, not after it’s deployed.

I’m glad you defragged it, rather than fragged it…

The underlying issue with an LLM is that there is no “learning”. The model itself doesn’t dynamically change whilst it’s being used.

This article sets out a process that gives the ability to alter the model, by “dialling up” (or down) concepts. In other words, it’s changing the balance of the weight of concepts across the whole model.

Altering one concept is hardly “learning”, especially since it’s being done externally by researchers, but it’s a start.

A much larger problem is that the energy consumption is several orders of magnitude larger than that of our brain. I’m not convinced that we have enough energy to make a standalone “AI”.

What machine learning actually gave us is the ability to automatically improve a digital model of things, like weather prediction, something that took hours on a supercomputer to give you a week of forecast, now can be achieved on a laptop in minutes with a much longer range and accuracy. Machine learning made that possible.

An LLM is attempting the same thing with human language. It’s tantalising, but ultimately I think the idea applied to language to create “AI” is doomed.

It’s going to take some time. I’ve been there as have plenty of people who came to me for support when it happened to them.

While right now you’re thinking of it in terms of loss, you can also celebrate the lightness that comes from not having the data anymore.

There’s more…

What was the funniest thing you remember that was in there?

Now consider that you remember it. You don’t need to check, you remember the things that made that memory funny.

So, take a deep breath, add it to the list of stupid things you’ve done to date that didn’t kill you and then go and drink a glass of water and go for a walk.

This too will pass.

Manufacturer warranty?

As opposed to the real apps that … steal your data?

I don’t know, but I doubt that the frequency response of a mobile phone microphone is either linear or consistent across sound level.

I don’t even think you could compare two sounds with different frequencies, but I don’t know.

I suspect that calibration of any such thing would require a whole lot of infrastructure, consider for example the angle of the phone in relation to sound and the impact of holding the phone in how it affects vibration and noise damping.

You might be able to use a calibrated sound level meter and pair it via Bluetooth with your phone, but I think that’s going to be as close as you might get.

In the past I’ve tried a wired USB microphone, but the OS isn’t real-time, so the jitter was horrendous. A pi would give you a more consistent result.

Note that there is no calibration of audio hardware, so the level of usefulness of any such software would be strictly limited.

You can change how long a phone rings for. Talk to her telco for both landline and mobile.

In my experience, if someone doesn’t want to answer the phone, strapping it to their arm is unlikely to make any difference and in my experience they’re more likely than not to leave it on the charger.

Long battery life and tiny battery are on opposite ends of physics. Pick your poison.

Health monitoring is unlikely to be transmitted to emergency services, except iOS fall detection.

iOS and Android are both tracking as much as they can get away with.

Remote management is likely only with devices used in corporate settings.

Looks oversized to me, that’s 3.25 cm x 3.25 cm, looks like they could take it down by 40% and still call it a 2x2 😇

It turns out that some clients don’t show my name together with my account. My name is Onno.

Thank you, that’s not something I knew, the three clients I’ve used show both the account and the name.

Edit: This is weird, my current client (Connect) shows the name, but only for my account, not for any other account.