Posts by captain veg 2920 publicly visible posts • joined 12 Jun 2009
When I was a nipper living in mid-Kent it was always in the bushes or brambles near the main road that you found some kind of porn magazine. I never knew why. We -- me and my equally juvenile mates -- found them gigglingly amusing.
Are we now to see equally baffling laptops apparently abandoned in the roadside foliage?
-A.
"A block list is a good thing."
Is it?
It's debatable at best.
If someone maintained such a list that was guaranteed not to contain any false positives or collateral damage and was kept up to date *very* regularly, then maybe. But it still doesn't address the actual problem, which is that this crap gets sent in the first place, wasting internet resources which we all end up paying for. I care much less that it is blocked from being delivered into someone's mailbox.
On the odd occasion that I get spammed I report it to the upstream providers and anyone else I can identify as having the service abused by these scumbags. This is the correct response, not putting your fingers in your ears and shouting I CAN'T HEAR YOU!
-A.
I was seriously fearful, back when Excel first incorporated VBA as an alternative to the old skool macros, that my team would be required to support these putative know-nothing coders.
It didn't happen. Not even the slightest hint.
I have no fear at all that ordinary users will start writing code nor even paste in LLM-generated slop. It would be like expecting them to write poetry instead of Teams agendae.
-A.
"This experience helps users quickly understand content, take action with fewer steps, and get more value from Copilot while extending productive browsing time in Edge."
Right.
In reverse:
I don't want to use Edge at all, but if I did I wouldn't consider "productivity" any kind of metric.
I find no value whatsoever in Copilot.
I understand perfectly the content of the pages I browse because I am a human being and I chose to read that content.
Thanks, and fuck off.
-A.
> I pretty much just read it aloud in my head and once I'd internalised the accent, it just made sense for me.
Well, bully for you. I stand by "impenetrable".
In the real world I live in France and spend a lot of time in Spain and those parts of Iberia where Catalan is spoken and so, frankly, I have more important living languages to worry about. Quite a lot of my interlocutors can hold a conversation in modern English, which is unfair but useful, but I would estimate that approximately none of them would cope with even slightly archaic English. Why should they?
-A.
> our bally language and we invented it 1000 years before you chaps decided to go your own way.
I had no idea that you were a scholar of old English.
Personally, like most Brits, I find Shakespeare fairly hard going, and America was already a thing by then. Chaucer? Pretty much impenetrable, though the swearing is fun. Back in the 770s "our" language was, er, a totally unintelligible hodge-podge of germanic and norse dialects..
-A.
Budget PC manufacturers would do well to simply put less expensive RAM* in the box and ship it with a less gluttonous operating system. I seem to remember this was tried a few years ago with sufficient success that Microsoft had to stay the execution of Windows XP rather than risk consumers realising that Linux could do the job perfectly well at zero cost.
-A.
* So long as it is user-upgradeable, of course.
> Facebook could easily identify at least 50% of those people who smugly announced they were staying under the radar by not signing up
Er, what?
I haven't "signed up". How exactly do they identify me?
I ought to mention that I never smugly announced the fact (whatever that means). Does that protect me from this dark magic?
-A.
Quite a long time ago, the last time I travelled through Gatwick our flight was up on the monitors for, approximately, ever, simply displaying the non-useful "on time". At the departure time this changed to "last boarding", resulting in a mass stampede through the terminal.
I don't fly any more.
-A.
Thank you for your offer to clarify.
We're talking about SMTP here. There is no security. Nor are there any labels like "Confidential" let alone COMMERCIAL IN CONFIDENCE, with or without the block caps.
I'm sending through my own MTA, which I trust as much as any other computer system, but rather more than any over which I have no control.
Sorry, still don't get it.
-A.
> I don't know what you're going to do with my email when you get it, and I have no way of knowing.
Yep. This is why I initially refused to correspond with any Gmail victims, back when they felt it necessary to spell out the fact that the content would be used to "personalise" the poor sods' advertising exposure. That principle was correct, but eventually it meant that many of my friends and relatives were cut off.
I'd like to think that they, the victims, are the only losers here. But is it a mistake to send anything at all into the Google maw?
Yes.
Does EU data protection law worry them in the slightest?
No.
-A.
My first real exposure to a computer was also a DEC machine, known as the "ten", and accessed over an acoustic coupler. Unlike the author it never occurred to me that I could run some kind of game on it, and my first basic program converted temperatures between fahrenheit and celsius.
At the time this was quite valuable. My generation had been brought up on fahrenheit and didn't get celsius (which we called "centigrade"), so this very simple program was seriously useful.
So I never saw the machine as a gaming platform, and so not a means of entertainment, and so weird AI interactions don't impress me much.
-A.
I bought a book last year on running LLMs with Python.
In the very first code example the author asks the model to write a password generator in Python, specifying precise complexity rules.
At the time I was just starting in Python, but even to me the code seemed odd. The author didn't seem to have noticed that it couldn't run at all due to illegal type conversions, and, even if you fixed that by hand, the output did not in any way observe the complexity rules. It was, in fact, superficially plausible but completely wrong.
Rather than actually test it, the author simply asked the same LLM to examine the code and describe what it did. It regurgitated more or less the original specification. This was presented as proof of the code's high quality.
I didn't read the rest of the book.
-A.
I don't get why this is so hard to understand.
> Imagine using an AI to sort through your prescriptions and medical information, asking it if it saved that data for future conversations, and then watching it claim it had even if it couldn't.
The LLM simply synthesizes a plausible response to your prompt based on a statistical model derived from its training data. There is nothing else. Each response is an extrapolation from your prior prompt.. You can ask it whatever you like, but this will always be true.
Why is this so hard to understand?
-A.
Bob, I'd forgotten .inf files.
Unable to magic up the secret incantation to make my OEM CRT monitor work with any kind of Linux distribution that I tried I resorted to asking the manufacturer (or, at least, their European distributor). I actually asked, if I recall correctly, for the necessary parameters to get my Xorg config working. They proudly sent me a "device driver" file, which turned out to be a wholly useless Windows .inf.
I ended up buying a new monitor. My first flat panel.
-A.
"Outsourcer tells MPs AI is prioritizing cases as thousands of civil servants face delays"
I suppose that when you have no actual intelligence then you'll take whatever kind you can get. However...
"Capita is banking on Microsoft Copilot to help rescue the backlog of cases it has inherited"
I didn't realise you could bid for an inheritance.
Craproilite wanted that business and competed, somehow successfully, to get it. Bitching about your winnings after the fact seems ungrateful at best.
> Chris Clements, managing director of Capita Public Services, said Microsoft's AI service was scanning initial contact forms and helping the outsourcer examine case documents
I do hope that the data subjects all freely consented to having their PII snaffled in this way.
> Copilot automatically reads and understands
Already stated above, no, it really doesn't. If you believe that then you shouldn't be allowed anywhere near the project. Or any project.
-A.
"while MATE isn't quite as light as Xfce, it does run perfectly well on computers lacking full hardware 3D acceleration"
I'm happy with Cinnamon on my daily driver and Mate does the job very well on really low-spec gear, the kind you can pick up for a hundred quid brand new. However, there's one scenario where even Mate slows things down enough to be annoying, and that's running remotely with XRDP. There Xfce is noticeably snappier.
-A.
Yes.
Absolutely.
Completely.
I would go as far as to say that anyone who can't understand some kind of assembly is not really a programmer.
Still, the answer to the conundrum at the heart of the article is simply to place the boolean result into some storage location and to return it after a fixed delay on a timer. Isn't it?
-A.
'OpenAI argued [PDF], among other things, that "users [of ChatGPT] were warned 'the system may occasionally generate misleading or incorrect information and produce offensive content. It is not intended to give advice.'"'
And, of course, in no way at all have you been presenting it as such.
Cunts.
-A.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2026
