andrewducker | Interesting Links for 23-05-2025

Entry tags:

Interesting Links for 23-05-2025

1. Why are Scotland's councils so short of cash when tax is going up?: (tags:scotland tax )
2. Greenland Signs Lucrative Minerals Deal with Europe in Blow to Trump: (tags:Europe USA materials Greenland trade )
3. How I Beat NES Mario in 0.000005 Seconds (the nerdiest video I have ever watched. If you've seen something nerdier then do let me know): (tags:video games programming technology mario )
4. Trump's image of dead 'white farmers' came from Reuters footage in Congo, not South Africa: (tags:USA politics Africa southafrica )
5. Programmers spend 5% of their time editing code, the rest is mostly understanding it/the issue they're trying to solve..: (tags:programming research )
6. Which word made you old? (Mine was GOAT): (tags:language age change comic )
7. Explaining "what a species is" turns out to be very very tricky: (tags:video ontology life )

#6: I felt old just a couple of weeks ago when someone responded to one of my Mastodon posts by saying "That's so based!"

I looked it up. Apparently it has a wider meaning of "well done for doing a thing the speaker approves of in spite of the thing also being controversial or unpopular", and an alt-right narrower meaning of "well done for opposing wokery" in particular.

In my case the comment was a response to a software tool I recently wrote to help people with anomalous colour vision, i.e. specifically intended as a disability-ish aid. (It's general enough to have other uses too, but that was the use that I particularly intended it for.) That might well make it the single most explicitly woke program I've ever written, so I have to assume the commenter didn't have the alt-right meaning in mind!

(The program has a certain bodge-nature, which I assume is the controversial-ness the commenter had in mind.)

Edited 2025-05-23 11:25 (UTC)

I would have said, “Based on what?”

Based has come to mean "A moral thing that you should be respected for doing." - basically the kind of thing that you could be referred to as a mensch for doing. At least in my experience.

And they were right, an awesome thing to do.

a software tool I recently wrote to help people with anomalous colour vision

My colour vision is anomalous (I have weak red cones). How can I find more about this tool ?

The toot where I announced it.

The inspiration was that a friend with anomalous colour vision said, elsewhere on Mastodon, "I find red on black text hard to read, but authors of colourful terminal software use it a lot. It would be nice if I could change it to be white on red, which I can read easily".

They didn't want to do this by hacking on their actual terminal program (which is perhaps the most obvious approach, at least if you're a programmer), because carrying a downstream patch to a program like xterm is a pain, especially if the program is constantly receiving updates from your distro.

I had the idea that you could do the same job by means of a separate program that sits between xterm and your shell session and just rewrites colour-setting control sequences, eliminating the problem of merging two sets of changes to the same source code. So I threw one together, and there it is (in a pointless demo mode where it just changes the colours). It has the key feature that it can think about both the foreground and background colours together, and rewrite them as a pair, so that it could map "red on black" into "white on red" as specified.

But at the moment, it's a starting point for experimentation: it doesn't have a built-in set of colour translations intended for any given type of unusual colour vision, and it also doesn't have a convenient configuration language. Anyone wanting to try it out at the moment will have to edit the C source code and recompile (which I know the friend in question is capable of). If it turns out to be useful in the long term then maybe I can change that, but I'd want to get some input from potential users first, because I don't even know what kind of configuration language would be convenient.

My Gen Z socially-conscious progressive kids use "based" a lot.

I too would be curious about your tool! I work for an organization that encourages disability inclusion and educates educators accordingly as well as providing direct services to students. They mostly use Chromebooks and tablets with students, though, as that's either what the school uses or what the students have for communication apps. In the office we have to use Microsoft Windows.

Then you might be disappointed to find that it's specific to Unix-style terminals, and can't do anything with a GUI, where there's no similarly narrow channel that all colour decisions go through.

(But I suppose it might be possible to run it in a Windows console, now that those are – or can be – mediated through the same system of escape codes. Hmmm.)

#5: many years ago, in one of my university vacation jobs, I had a conversation with my manager about input devices, in which I said I preferred to use the keyboard to the mouse because (if you can type well) you can input data at a higher rate, and get things done quicker.

My manager's response was something like "You should be neither typing nor mousing that fast". He felt that if your input speed is the limiting factor in working with a computer you're doing it wrong – you should be thinking more and inputting less.

I disagreed with him then, and I still disagree with him now. But I don't disbelieve your statistic of only 5% time spent editing code. (In fact at work it might be even less than that, for me – both editing code and understanding the problem have to share my work time with meetings, bug-tracker paperwork, etc…)

But the thing is that, for me, "understanding the issue I'm trying to solve" also involves a lot of input, because I very highly prioritise hands-on investigation. Much more so than a lot of my colleagues, who are constantly looking at an error message and then just scratching their head trying to work out what might have gone wrong. If I can't see it immediately, I'll try to get more information, by typing more commands – running the program under strace, or under a debugger, or filling it with diagnostics and re-running, or running it differently, or cutting down the input file, or something which will give me more clues than I already have. So my speed of input is still important even though I'm mostly trying to understand the problem, because the way I understand the problem is highly interactive!

… perhaps a nice analogy is that the colleagues I mention are treating debugging like Sudoku: stare at the limited set of clues you have, and think harder and harder until you manage to deduce something from them. Whereas I'm treating it like Minesweeper: you're almost never expected to be able to solve the whole thing from just the clues you're currently looking at, and 99% of your play is focused on finding a safe way to collect additional clues.

Minesweeper? I don’t have that on my computer anymore, but I used to win at it about 19 times out of 20.

Which size ?

For about the last 30 years, I have played it (and/or the unix implementation xdemineur) without using the flags, so my scores are much lower.

I think it came in three sizes.I played the middle size.

I also played it without flags. I found it both easier and more fun this way, plus it was neat seeing all the flags suddenly appearing when I won the game.

Definitely. I want to get a clearer view of the thing that behaves unexpectedly so I need to do things to get that view.

Also I simply find text-based interfaces less annoying and confusing than GUIs. I don't want to try to be aiming my clicks precisely and suchlike when I just want to get something done. (And goodness, modern UIs where things pop up unexpectedly based on hovers and such.)

Hate GUIs. Hate. Hate. Have hated since Windows was invented.

I'm with you on wanting to work my way through, check state at various points in a debugger, etc. I like hands-on tweaking.

I like hands on. But my input speed into a computer by any means is incredibly slow. I never learned to type, not in 30+ years of daily keyboard work and a few early deliberate attempts.

My READING comprehension speed however, has always been insanely fast. So, I guess you win some you lose some

Those aren't incompatible if your manager had thought about it. I type quickly and this helps me with those bursts of "ok I know what to do here and let me get it out Right Now". That doesn't mean I spend most of the day that way. I'm guessing you're in a similar position.

And I'm with you on "to understand it I need to interact with it". Yes we also need to stare at code, but it's too easy to make assumptions while reading that a little actual poking would quickly dispel.

"Yeet" was my old word, but I've found it very useful. (The best word I can think of that means the same thing is "huck" but I've only seen that in books.)

I felt like I'd entered an alternate reality when somebody said, of another friend, that she was going to "yeet the yute", which is the strangest way to talk about a hysterectomy that I've ever heard. (I had not heard "yeet" before and had to look it up.)

I mean that at least has the ring of deliberate word play. That's not a thing you'd say because you think it's the clearest expression of your meaning (surely?), it's because it's the silliest. :-)

I assume the alliteration was the goal, yes, and if I'd known the word "yeet" at the time I probably would have laughed. :-)

My word was, "rad."

Golly! That's an old one!

As is "golly" (I see what you did there.)

Golly goes back to 1775, so I reckon that unless I have a vampire or a ghost reading this I'm probably safe!

On #5: THIS. The exact numbers aside, this is why I consider AI-based coding tools to not just be problematic in terms of their quality -- more importantly, they're solving the wrong problem.

Writing code shouldn't be the thing that needs automation; frankly, if it is, I generally think the codebase isn't well-enough factored.

What I find LLMs useful for is simply as a glorified search engine. That's not a small thing: my experience over the past six months is that phrasing my question as a question in Kagi (which turns on the Quick Answer feature), and then double-checking the citations, takes me about half as much time as a conventional search does. Given how much of my time is spent on simply looking stuff up, that's not a small thing.

But code automation? Until and unless that produces truly high-quality code, without dumb bugs, based specifically on my codebase, it's taking the minor problem and making it worse, which is a foolish waste of my time.

This is what I wrote on Andrew's FB post sharing this link...

"A (former) botanist writes... This is because "species" is a human concept, born of the very human need to classify things. Individual or populations of other organisms neither know nor care what species they are.

When I studied bracken, it was classified as a single species with two subspecies and twelve varieties. A couple of years after I'd finished working with it, my former supervisor concluded it was in fact twelve different species. According to Wikipedia it is currently considered to consist of one or two species - depending on which species the article you read is writing about! (My supervisor was undoubtedly wrong: he was very much a splitter, seeing subspecies all over the place, whilst I am very much a lumper!)

DNA studies can tell one how closely related a group of organisms is and the relationships between them, but the definition of a species and where the break between one species and its closest relatives is always subjective.

And it the creator of that video thinks plants are difficult (they aren't!), wait till he finds out about fungi and bacteria...!"

Thank you!

Flat | Top-Level Comments Only

Interesting Links for 23-05-2025

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

No such thing as a species (or "it's complicated)...

Re: No such thing as a species (or "it's complicated)...