All of these Long Links pieces have begun with more or less the same words, so why stop now? This is an annotated parade of links to long-form pieces. Most people won’t have the time (nor the weird assortment of interests) to consume them all, but I hope that most readers will find one or two reward a visit.

Radisson (and Groseilliers) · I don’t know if it is still the case, but in my youth, Canadian elementary education included several overexcited units about the Coureurs des bois, early European settlers in “New France” (now Québec) who ventured, by foot and canoe, far to the north and west, mostly engaged in trading with the indigenous peoples: trinkets (and later, serious hardware including guns) for furs.

The names I remembered were Radisson and Groseilliers, but I don’t recall learning much about who they were and what they did. Then I ran across the 2019 book Bush Runner: The Adventures of Pierre-Esprit Radisson and, wow… The writing is pedestrian but who cares because what a story! Radisson lived an absolutely astonishing life. He went as deep into the bush as anyone of his era, interacted intensely with the indigenous people as business partner, friend, and foe, worked for Charles of England and Louis of France (changing sides several times), in 1670 founded the Hudson’s Bay Company (recently, 355 years later, deceased), and fortunately took notes, a copy of which was preserved by Samuel Pepys.

I learned more from this book’s pages about the early history of Upper and Lower Canada than all those elementary-school units had to offer, and had loads of fun doing so. I guess this is a fairly Canadian-specific Long Link, but I think anyone interested in the early history of Europeans in North America would find much to enjoy.

Music · It’s rare these days that I discover interesting new musicians, but here are two of those rarities.

Lucie Horsch plays recorder, you know, the cheap plastic thing they use to introduce second-graders to music. It’s actually a lovely instrument and I wish we would switch to its German name, “Blockflöte”, which to my ear sounds a bit like the instrument does. Anyhow, check out this YouTube entitled only Lucie Horsch - Bach, annoyingly omitting any mention of which Bach. Annoyance aside, it’s a pretty great performance, Ms Horsch is the real deal, full of virtuosity and grace.

I got an unusual mid-week message from Qobuz, all excited about The New Eves’ new record The New Eve Is Rising. So I played it in the car on a long crosstown drive and now I’m all excited too. The New Eves are talented, musically surprising, and above all, insanely brave.

Their music doesn’t sound like anything else and flies in the face of all conventional wisdom concerning popular music. They take absurd chances and yeah, the album has klunkers amid the bangers, but when I got to its end I went back and started at the beginning again. I found myself smiling ear-to-ear over and over. Maybe I’m being a bit over-the-top here, but check them out: Mother is live. Cow Song is off the new album and strong albeit with forgettable video.

Life online · Every Long Links has hardcore-geek threads and there is no harder core imaginable than Filippo Valsordi’s Go Assembly Mutation Testing. I have always admired (but never actually used) mutation testing, and Filippo offers a convincing argument that it moves catching certain classes of bug from nearly impossible to pretty easy. Good stuff!

And of course we can’t ignore genAI and programming. Most of you are likely aware of Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity, but I’m linking again to boost its visibility, because hard quantitative research on methodology is damn rare in our profession. I will confess to being a little (but just a little) surprised at the conclusions.

It is apparently quite possible that Intel will exit the business of making high-end chips, leaving TSMC with a global monopoly: Intel and the Wide Open Barn Doors. This is an unsettling prospect. Not, I have to say, surprising though. I’ve sneered at Intel leadership cluelessness for years and years, see here and here.

Finally, here’s the charmingly-titled How to Surf the Web in 2025, and Why You Should. I love this piece.

Class Reductionism · The news keeps making me want to build something around the classreductionist.org domain name I’ve owned for years.

The tl;dr on Class Reductionism is something like “In the best possible world it’ll take generations to disassemble the global tangle of intersectional oppression, but we could treat the symptoms effectively right now this year by sending money to the poor. I’m talking about Universal Basic Income or suchlike. I wrote a couple thousand words on the subject back in 2023, and there are complexities, and I probably won’t put up that site. But I still do maintain that a very high proportion of our societal pain is rooted in the egregious inequality, and consequent poverty, that seems a baked-in feature of Late Capitalism.

Let’s start with Nobelist Paul Krugman, who’s been writing an “Understanding Inequality” series on his paywalled newsletter and then republishing a gratis version, start here. Very data-dense and educational. Hmm, that site is slow; there’s a livelier table of contents here.

Don’t kid yourself that this is just an American problem, see ‘The Better Life Is Out of Reach’: The Chinese Dream Is Slipping Away.

Let’s pull the impersonal veil of facts and figures aside and focus on the human experience of what we used to call Class Struggle. Confessions of the Working Poor is beautifully written and opened my eyes to lifestyle choices that I didn’t even know some people have to make.

But hey, there are people who are just fine with this: Delta's premium play is taking advantage of the growing economic split.

Look, being class-determinist-adjacent doesn’t mean you should ignore intersectional awfulness: What We Miss When We Talk About the Racial Wealth Gap.

No more sections · The remaining Long Links refused to be organized so I had to turn them loose; call it the Long Tail.

The Venetian origins of roman type. You might think you don’t care about typography but still enjoy the pictures and descriptions here.

This guy is a full-time Coyote researcher. What a great gig! I’m an admirer of those animals and how they’ve carved themselves a comfy niche in most of North America’s big cities. (Even if it means that you better not let your cat out at night.) They’re also remarkably attractive.

Here’s another long list of Long Links, and many of you will wonder why anyone would choose to browse it: The Best Camera Stores in Tokyo: The Ultimate Guide. Some of the interiors are remarkable.

Oh, while we’re on the subject of photography: A Photojournalist Took a Fujifilm Instax Camera to a Mexican Cartel Wedding.

GLP-1’s (i.e. Ozempic and friends) would probably dominate a large section of the news if weren’t for all the political craziness. Here’s one small example: How GLP-1s Are Breaking Life Insurance.

Science is hard. There are lots of largely-unsolved areas, and “gap-map.org” tries to organize them: Fundamental Development Gap Map v1.0. The UI is a little klunky but the thing still sucked me right in.

I’m going to give the last word to Laurie Penny. I don’t know what we’d do without her. In a time of monsters: do we have any ideas for surviving the zombie apocalypse that aren’t nightmare patriarchy?



Contributions

Comment feed for ongoing:Comments feed

From: Ryan Baker (Aug 10 2025, at 06:46)

On the GenAI efficiency study, I'd pay close attention to this caveat toward the end of the study "It also may be the case that there are strong learning effects for AI tools like Cursor that only appear after several hundred hours of usage—our developers typically only use Cursor for a few dozen hours before and during the study."

While it would be hard to say for certain without additional studies, it seems highly plausible that this plays a key role that could reverse the findings entirely.

The developers were experienced, so thousands of hours of experience in one model, vs. dozens in the other. But it's probably a bit deeper than that too. If I was given a "challenge" and had a bit of experience with a new tool, the way I'd tend to approach it would be to push myself to try new techniques. Paradoxically, that though would often not produce a faster result. The fastest result would almost always be to use whatever techniques I had already had mastered, and not experiment. But being the continuous improvement minded person I am, it would be very hard not to be drawn toward experimentation. Probably the only solution would be hundreds of hours of experience.

I also think in the end, studying this will be harder than expected. While it's tempting to compare individual to individual, it's hard to imagine that any individual really keeps two separate versions of themselves, GenAI enabled and not. Thus, if you did include any developers with thousands of hours of experience with GenAI tools, and ask them to perform a task without them, they'll probably do worse than they would have done before they acquired those thousands of hours of experience.

[link]

author · Dad
colophon · rights
picture of the day
August 04, 2025
· The World (157 more)

By .

The opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.

I’m on Mastodon!