cesaref

Back in the 80s, as the home computer revolution got going, computers were typically wired up to small, cheap, portable TVs as a display device. These TVs used shadow masks, and the computer video output was typically modulated to a TV signal, and the TV was 'tuned' to the computer. All of this added large amounts of blur and distortion even before the signal was displayed on the TV.

By the mid 80s, it was maybe more typical to buy a dedicated CRT monitor, and the computer connected via composite, or maybe even an RGB feed to the monitor, allowing higher resolution and much improved quality.

For the well healed, this route also led to the holy grail, a trinitron tube!

At each of these changes, the aesthetic of the display technology changed, but probably the best memories come from the original blurry stuff as the magical moment of actually getting something out of a home computer.

brew install grep?

There are exceptions - here in the UK we have the RI: https://www.rigb.org/

The Christmas lectures are probably the most famous thing they do, and these have definitely moved in a more 'child' focussed direction. If you were attending the Christmas lectures in the 1850s however, the audience would have been middle class victorioans, and you'd have had Michael Faraday telling you about electricity, forces, chemistry etc.

https://en.wikipedia.org/wiki/Royal_Institution_Christmas_Le...

I would recommend attending one of their lectures if you happen to find yourself in London, just to be in the building, and to sit in the lecture theatre!

They are vitamin C to the mixture, which is actually what is developing the film. There's a commercial developer (Xtol) made by Kodak which is I believe vitamin C based, and it's a very good developer.

It would be interesting to know if the caffeine in the red bull is making a significant contribution I guess, otherwise the jagerbomb isn't really doing that much I think.

>so how do you get a privately owned train car and get it to the tracks or etc?

I think you wait in a remote bit of Nevada for a train to pass, and trigger a rock fall which causes the driver to slam on the brakes and bring the train to a stop just short of the rockfall.

Then, you and your posse jump out from behind some rocks and fire your revolvers in the air, and the driver sticks his hands up. There's much celebration, and back slapping as you discover the train also happens to have a massive amount of gold bullion on board.

The rest is a bit blurry, can't remember seeing what you then do, but it probably involves filing down the serial numbers on the frame or something like that?

Yeah, with those early machines it was inevitably about lookup tables, and producing the illusion of doing work which was otherwise impossible in the number of cpu cycles you had available. I did lots of this sort of thing for the C64, where similarly CPU cycles were short, and then you end up scratching around trying to find enough spare memory...

It's great to see people still exploring and pushing the boundaries of what these machines are capable of. Coding this on a more modern machine, getting the effect right, then back-porting to the Spectrum seems like a smart move and would have been a game changer during the prime of these machines!

If he'd been talking about the Acorn Archimedes, he'd have had a point.

This situation happens with plenty of compilers for plenty of languages. It's just software after all, we don't expect any software to be perfect, so why expect compilers to be perfect?

As for deprecating old C++, yes, I can see the appeal, but no, the community hasn't gone that route. C++ has always been a 'mix and match' approach, with different parts of the community using more or less of the language features, with the intention being that you can choose what to adopt when, and move older code bases forward (or not!) depending on individual concerns.

It's one of the C++ strengths, although it doesn't always feel like it.

And yes, people being able to use your code for whatever they want is absolutely more open than having restrictions on how/who gets to use it.

One other model that can also work well is to dual license as GPL + commercial, so people who want to publish their work can use the GPL license but you can potentially fund the project from license sales to closed source users using the commercial licensing option. I see this a fair bit in the audio community I work within.

I remember these discussions in the 80s...

The problem as I see it is that LLMs behave like bratty teenagers, believing any old rubbish they are told or read. However, their voice is that of a friendly and well meaning adult. If their voice was more in line with their 'age' then I think we'd treat their suggestions with the correct degree of scepticism.

Anyhow, overall this is an unsurprising result. I read it as 'LLMs trained on contents of internet regurgitate contents of internet'. Now that i'm thinking about it, i'd quite like to have an LLM trained on Pliny's encyclopedia, which would give a really interesting take on lots of questions. Anyone got a spare million dollars of compute time?

I was involved in writing ATC software in the 90s for a European country (the FATMI system for Finland), and they were definitely using paper strips at the time, and I believe the design did not change this. I wasn't involved in the flight strip printing though, I was working on SIDs and STARs, airspace sectorisation, that sort of thing.

It would be interesting to know how things have changed since then, as obviously nearly 30 years has passed since that system would have been commissioned!

Generally speaking, the db scale is very useful for many practical situations, and this is overlooked in this critique.

As have been pointed out, it's just a power ratio on a logarithmic scale, but this has many benefits, the main one being that chaining gain/attenuation in a system is just a case of adding the db values together. 'We're loosing 4db in this cable, and the gain through this amp is 6db, so the output is 2db hotter than the input'. Talk to any sound engineer and you'll do this sort of thing successfully without necessarily understanding the science, so that's a massive win.

If you look at professional distributed audio systems (Dante, AES67 etc) you'll find that they all require PTP support on the hardware to achieve the required timing accuracy, so yes, you need <1ms to get to the point of being considered suitable if you are doing anything which involves, say, mixing multiple streams together and avoiding phasing type effects.

However, it very much depends on what your expectations are, and how critical your listening is. If no one is looking for problems, it can be made to work well enough.

The problem is that electrolytic capacitors can leak, even if not used. So, you can store a perfectly good computer and retrieve a ruined motherboard a number of years later. SMD capacitors are bad for this, the MZ80K predates all of that though, and I think only has a few through hole capacitors on the mainboard.

The power supply should also be checked, and in this case, it's safety capacitors which are the risk there, since some of the older ones (RIFAs for example) tend to be hygroscopic, and if they crack they end up blowing next time mains is applied across them. I've not had any problems with power supplies killing the machine with an over voltage, but i've heard this is also a risk.

Anyhow, 45 years or so is time for a recap, and it'll do another 30 years before it's time again...

Sounds like you are either very unlucky, or have a useage pattern (or install problem) that is triggering issues. I'd definitely try and work out why they failed before pulling the trigger on another machine, no matter what the make, unless of course the failures are obviously defects, in which case i'd be complaining to Miele and seeing what they can do about it.

Here in europe we have Miele who make various white goods, and their washing machines are designed to actually last, rather than fail once the guarantee has run out.

I think they plan for 20+ years lifetime. Our one is getting on for 18 years old now, and has had no issues, or for that matter, any indication of problems.

https://www.mieleusa.com/e/washing-machines-1015696-c

The question is about classical (newtonian) physics, which is a mathematical model, rather than the real world. The question is asking when is there non-determinism in the newton mathematical model.

In the model, you have perfect spheres...

There's also the confusion as to whether people mean 'Greater London' or 'The City of London'. For the centre of the City, i've generally thought that Leadenhall Market is the centre, given it is built on the original Roman Forum.

For non-Londoners, the city was originally a walled city, and lies to the east end of what is now considered greater london. It's these days synonymous with the financial industry. There are special laws for it, honours like Freedom of the City, it's quite an interesting place.

Although the wall is long gone, there are place names which refer to it, and the gates which exited through it. So we have roads like 'London Wall' and locations like Bishopsgate, Aldgate etc. Newgate was added in the 12th century, so not exactly 'New' these days, so not a very future proof naming convention...

https://leadenhallmarket.co.uk/history-of-leadenhall-market/

https://inspiringcity.com/2013/04/13/the-seven-gates-of-lond...

I can certainly relate to this experience. I remember when it was introduced, I was very wary of getting too much 'auto' into the codebase, certainly as the team could be a bit 'gung-ho' adopting stuff just because it was there.

However, in hindsight, I think I was being overly conservative, and it worked out well, and adoption didn't cause any obvious problems.

Your concerns about learning a new library are valid, but the problem is the library if it's not clear, or well documented. To lay responsibility for this at the door of auto is a stretch. You can write great and terrible code with a number of language features (dubious use of goto is the classic example), and it sounds like you are tackling a library which could do with some love, either in documentation, or to clarify it's conventions.

Let's consider the original list. It starts with DSLs. Are we saying that using SQL 'almost never works'? Shader languages? BPF? If I squint a bit i'd say that all the inference engines are either DSLs or APIs which wrap an underlying DSL, so TensorFlow etc.

I could probably come up with similar examples from the rest of the list. If the message is 'system ideas which seem simple but are hard, so you should only use existing off the shelf examples of' then i'd certainly have more sympathy for the statement.

It's a bad example. Lots of finance firms use AI to manage their financial affairs - go and investigate what is currently considered state of the art for trading algorithms.

Now if you substituted something safety critical instead, say, running a nuclear power station, or my favourite currently in use example, self driving cars, then yes, you should be scared.

Sounds like CS is not for them, and they find something else to do which is more applicable to their skills and interest. This is good. I don't think you should see a high drop out rate from a course as necessarily indicating a problem.

Agreed, there needs to be trust, and how this used to work was that the photographer would gain a reputation for their accurate portrayal and would be picked up by an agency (e.g. magnum). They would supply the picture editors for the newspapers, so the public would have a chain of trust back from the the paper to the photographer.

The equivalent for digital photography is that cameras can sign images - I believe there are some canon/nikon cameras which supported this to allow digital photos to be used in court, but I don't think this idea was widely adopted, so a typical snap on your phone doesn't provide any sort of trust, which is surprising really, but there we are, this is the world we live in.

The problem is broader than this. There is a tradition of using photography to record real world events (photojournalism) which has been quite broadly adopted within our cultures - we believe a photograph as reflecting events that have happened.

The problem with these AI generated 'realistic photojournalist' images is that they erode the belief in these images as being factual representations of real world events.

Just think of the impact that, say, war photography has had on the world, and our understanding of the horrors that our governments decisions have had on people's lives. Would Nick Ut's 'The terror of war' have had the impact it did if AI image generation was available at the time? I believe there was at the time an attempt to say it was staged, but there are many many ways of dismissing an image these days.

So, I think we have lost something, but it's much greater than the connection with a subject that the author is concerned with.

You mean, would it be possible to oversample the algorithm, run it effectively at a much higher sample rate and reduce the audibility of the aliases? Yes, it'll definitely help, but no, it's not possible to eliminate this effect, even with monstrous levels of oversampling.

The problem is that, say, a piano keyboard runs from 30Hz to 4Khz, so to get an even tone, at the top of the keyboard, you are only hearing 4 harmonics of the fundamental, whilst at the bottom, you potentially have roughly 1,000 harmonics in the audible spectrum.

Without fiddling with the voice with keyboard scaling, that same tone with 1,000 harmonics played at the top of the keyboard will generate a 4Mhz frequency, so you'd need a 2Mhz sample rate to avoid the aliases reflecting into the audible spectrum.

So that's 45x oversampling, which is a scary amount of effort to throw at the problem, to resolve the issue for this one contrived sound.

Now if you modify the above sound, and double the modulating frequency (the operator pitch) then you'll double the oversampling required, and you have basically limitless control to create higher and higher levels of harmonics to tame, so you can always fiddle with the setting to produce aliasing if you try hard enough :)

On the subject of the DAC used, and why it was 'only' 12 bit. I believe the instrument actually runs all of the voices through a single DAC, so it is basically 16x oversampled, and the voices are held by a sample and hold circuit for all 16 voices before they are summed to the analog output.

If the later DX mark II is anything like the TX81z (which is also said to have a 16 bit DAC) it actually uses a 16 bit dynamic range DAC, but with a 10 bit resolution (a YM3012 for the TX - https://www.alldatasheet.com/datasheet-pdf/pdf/1179489/YAMAH...)

This is definitely the age of interesting DACs, with all sorts of interesting noise profiles. We're really spoilt these days with the quality from delta/sigma DACs, thinking that this has always been the case!

There was mention of the sound becoming 'dull' at the higher octaves. This is not a limitation of the DAC as was suggested, but caused by the phase modulation algorithm aliasing badly at the top of the range. Basically the modulation causes higher harmonics to appear in the output, and if these are higher than nyquist (half the sampling rate) they reflect back into the audible range, and produce unpleasant very digital sounding artifacts.

So, the DX designers avoid this by adding keyboard scaling for the operators - as you play higher up the keyboard, sounds could reduce the amount of modulation applied, and hence reduce the level of the aliases till they were inaudible. Unfortunately if done too aggressively, this will lead to a noticeable reduction in the harmonic complexity of the tone, which is perceived as a dull tone.

'These numbers include all children of middle school age, not just those who attend school.'

I think the low scores are telling you more about the pressure on children in poorer countries to work to support their families, rather than the quality of teaching when they are able to attend school.

Yeah, if resolution is your thing, then you have unfortunately little option without spending proper money for medium format digital. It's worth it though, if you find yourself with the budget for a phase one or a fuji system!