Software: Weekly Summary (January 26 - February 01, 2026)

I kept poking through a week of short, punchy pieces on software and a few of them stuck in my head. They felt like little windows into how people are living with code right now — some practical, some wistful, some quietly angry. I would describe them as a mix of nostalgia, tinkering, and strategic squinting at where the business of software is heading. To me, it feels like the same old arguments are circling back with new costumes: AI, subscriptions, craftsmanship, and the perennial tug of usability.

A quick tour of the week

There are posts that are like a cup of tea you take by the window and read slowly. There are others that are quick espresso shots, straight to the point. If you want the exact pieces, there are some names to follow. For the historical angle you get John Paul Wohlscheid pulling an old interview with Mitch Kapor out of the attic. For the subscriptions and AI shuffle, Heather Meeker tells you where she dumped a few apps and why. Stephen Hackett nerds out on a Finder tweak that actually matters. Lucio Bragagnolo sings a small ode to TextEdit. alex wennerberg wants a revival of craft. And then there are practical guides and tips from Nacho Morató, Dave Pearson, Brian Fagioli and a few dark thoughts about attention economics from dead_neurons. Plus a useful podcast episode from Sherman On Software about rewrites and how they hide problems.

I’ll group the week into three threads that kept popping up: the human side of software (craft, small tools, TextEdit vibes), the economics and business of software (subscriptions, megaphone problem, Kapor’s old playbook), and the nitty gritty practical stuff (Finder tweaks, Linux distro, uv script, iTunes guide). There is overlap, sure. That overlap is where the interesting friction lives.

The human side: craft, small tools, and a love letter to simple apps

There was a soft chorus this week for the simple things. Lucio Bragagnolo writes in Italian about TextEdit and how its modesty can be a feature, not a bug. To me, it feels like someone telling you to slow down and stop chasing the giant Swiss Army knife of apps. TextEdit is basic. It opens, it saves, it doesn’t nag with an account to set up. It’s like grandma’s kitchen knife: humble, useful, and not trying to be everything at once.

That ties in neatly with alex wennerberg who worries aloud about AI turning online production into a race toward the bottom. He’s not just nostalgic. He wants craftsmanship back in software — the intentional, careful making that values longevity and clarity over metrics and churn. I’d say he’s painting a picture like the Arts and Crafts people did a century ago, but with code instead of furniture. He argues that as AI makes cheap stuff abundant, genuine care will become a rarer, more valuable thing. That sounds right on paper. It also makes me picture artisan bakers in a world of factory bread — fewer small shops, but those that stay feel special, and you notice the difference.

These two pieces push a common point: not every problem needs a cloud, an account, and a ten-dollar monthly fee. Sometimes the right tool is the one that gets out of your way. They repeat the same idea, and that repetition matters. It’s like when you tell someone a recipe twice because the second time you add the bit about how long the dough needs to rest.

The subscription shuffle and the promise of general AI tools

Heather Meeker lays out a personal experiment that feels familiar to anyone with a bulging subscription bill. She calls it the Great Subscription Pivot. She’s trimming specialized apps and moving into general-purpose AI tools that can handle a bunch of jobs — from receipt tracking to image generation. I would describe her shift as pragmatic: fewer single-purpose apps, more swiss-army-AI-tools.

But she also leaves a big question open: is this sustainable? To me, it reads like someone who swapped several single-purpose subscriptions for a single buffet pass. It’s cheaper and less cluttered, sure, but you start to worry about how long the buffet will keep serving the dishes you like. And then there’s the worry she mentions in passing: what happens when those general AI providers change price, limit features, or lock parts of the tool behind tiers? Suddenly the buffet becomes a la carte again.

That anxiety about sustainability ties to the larger business note in deadneurons. Their piece, which I’d call a reality check, says AI may cut engineering costs but it does not make attention cheaper. The cost of being heard — marketing, sales, user acquisition — stays high. This, deadneurons argues, will favor big players who can shout louder, rather than a thousand tiny SaaS shops. To me, it feels like standing at a fair where the loudest stall still gets the crowd, no matter how good the quieter vendor’s jam is.

Those two posts sort of speak to each other. Heather feels the convenience of consolidation; dead_neurons names the economic pressure that makes consolidation attractive and scary. The two together make a claim: convenience is winning, but it comes with a cost that isn’t always obvious until the lights go dim.

Old lessons in a new form: Mitch Kapor and the business of software

There’s an unusual slice of history in the week. John Paul Wohlscheid reprints a long ago interview with Mitch Kapor from 1985. You would think a 1985 interview wouldn’t feel relevant in 2026, but it does. Kapor was talking about customer satisfaction, differentiated services, and software piracy — the same axes people are squabbling about today.

It’s instructive to see how familiar some strategic concerns look when they’re thirty or forty years old. Companies still wrestle with product differentiation and customer trust. The tech stack has changed, but the worry about churn, piracy, and doing something customers actually want is ancient and stubborn. I’d say Kapor’s words are like a dusty road sign you find in the middle of a new neighborhood. The letters are faded, but they point the same way.

Kapor’s lens adds a small, useful dent in the hype. It forces a question: are we reinventing strategy every ten years, or just dressing the same problems in new jargon? That’s a small digression, but it feels relevant when everyone is promising that AI will change everything. Sometimes the big things stay the same.

The megaphone problem and why marketing still rules

Back to economics. The megaphone problem, again from dead_neurons, made me think of a market that’s part radio contest, part farmers market. You can build a great product, but if no one hears you, it won’t matter. The conclusion they draw is blunt: many products will never be sold to the masses. They’ll become internal tools, custom software, or quietly useful things inside companies.

This messes with the usual startup dream. You know the one: build a little team, write some beautiful code, scale to a million users, retire early. The article says the middle will get scrunched. You’ll have a few giants who can afford slamming the megaphone, and a long tail of internal apps or bespoke products. The tiny SaaS dream becomes harder. That felt like a punch of cold air.

It’s a good reminder. Building software is more than code. There’s a whole ecosystem: sales, support, trust, distribution. You can lower one cost with AI or tooling, but the rest stay put. That creates a structural bias toward consolidation. And when that happens, you don’t just change companies. You change culture.

Craft vs scale: AI, low-quality content, and finding value

The craftsmanship argument also comes up when people worry about AI flooding the internet with cheap stuff. That’s not just noise in music or blog posts. It’s code too. alex wennerberg warns that when platforms prize metrics over craft, quality gets squashed. The piece suggests that if AI makes plenty of acceptable code, then the few craftsmen who insist on higher standards will matter more.

That resonates with the Kapor piece in a strange way. In both cases, the market will ultimately reward whoever can deliver sustained value. The difference is that one says value is a product and features, the other says value is craft and trust. I’d say they’re both right and they’re talking past each other at the same time.

Rewrites, accountability, and the bugs that hide in plain sight

On a related but more tactical note, Sherman On Software has a podcast episode about rewrites and how they create accountability black holes. This is the kind of thing you only notice if you’ve inherited a codebase at 3 a.m. The episode explains why bugs survive rewrites: when teams split work, when nobody owns old behavior, or when the rewrite duplicates but doesn’t replace the original.

That echoed the craft conversation. Rewrites are often sold as a chance to do things cleaner, but they also create forgiveness for messy transitions and for ignoring legacy issues. It’s like moving house and leaving half your junk behind because you’re too tired to sort it. The bugs remain, and so does the debt. This is a practical warning. It’s also human. People avoid hard choices.

Small, useful changes: Finder columns, Linux Lite, uvx.sh

Now the week’s practical bits. Stephen Hackett wrote something that at first seems tiny — column auto-resizing in Finder. But if you live in Finder’s column view and wrestle with filenames that are too long or too short, this one change matters a lot. I’d say it’s one of those interface fixes you don’t notice until it’s fixed, then wonder how you ever lived without it. It reminded me of tidying a messy junk drawer: trivial, but satisfying.

On the operating system side, Brian Fagioli suggests Linux Lite 7.8 as a serious alternative for people fed up with Windows 11. The release tries to make life easy: cleaner interface, rewritten core apps, better hardware support. It’s pitched at people who just want their computer to behave. That’s a regional appeal, really — it feels like telling someone in the UK to try a proper cuppa instead of the supermarket swill. It’s pleasant, reliable, and doesn’t pretend to be exotic.

For developers who like to tinker, Dave Pearson had a small gem about uvx.sh and uv. He praises the uv tool for managing Python projects and highlights a helper site that simplifies installing tools without pre-install requirements. It’s the kind of piece that’s part how-to, part cautionary tale: convenient, but with a wink about the risks. That wink matters when someone suggests running shell commands from the internet. It’s like accepting candy from a stranger — tasty, but check the wrapper.

Manuals and guides: iTunes on Windows keeps on truckin

Then there’s Nacho Morató with a thorough guide on installing iTunes on Windows 11. Someone out there still needs the old all-in-one player. The guide is a reminder that software history lingers. Even when companies split a product into pieces, the old habits and workflows stay with people. iTunes is like vinyl collectors: plenty of folks moved on, but some refuse to give up what still works. The guide is practical and patient — the kind of thing I’d pass to someone who refuses to switch to streaming-only setups.

Small contradictions and repeated refrains

As I moved between these posts, a few contradictions kept popping up. People love convenience and consolidation. They also romanticize small tools and craft. People want AI to reduce costs, and they also fear that attention and distribution will keep costs high. The industry wants new tools to democratize creation, and at the same time the market concentrates power.

That kind of contradiction is human. It’s like wanting cheap beer on a Friday but also craving a corner pub where the bartender remembers your name. The two desires coexist and they tug in different directions. The blogging week reflects that tug.

A few small predictions and gentle nudges

You can sniff patterns in these posts. Expect more consolidation in the near term. Expect more talk about craft, and more pushback against commodified AI output. Expect user-facing UX fixes that feel small but make everyday work nicer. Expect to hear more about internal tools and bespoke software as public SaaS dreams dim a bit.

If you like tinkering, keep paying attention to the small guides. They’re the practical lifeboats — the Finder tweak, the Linux distro, the uv tool. If you like theory, watch the megaphone problem and the Kapor reprint. Both will tell you something important about where the money and attention are flowing.

Where to poke next

If one of these pieces nudges your curiosity, go read it. The Kapor interview feels like a time machine. The subscription pivot is a tidy experiment that might mirror your own cleanup. The Finder column fix is a tiny joy. The TextEdit piece is a gentle reminder to prefer simple tools sometimes. The craftsmanship piece is an argument worth chewing on, even if you disagree. The podcast on rewrites will make you think twice before promising a rewrite to stakeholders. And the Linux and iTunes guides are here for when you just want your machine to behave.

The week left a distinct taste. Part nostalgia, part practicality, part strategic realism. It didn’t hand out any easy answers. It did, though, collect a few sensible reminders: make things people can live with, be honest about costs beyond engineering, don’t forget that a small UX fix can mean more to users than a shiny new feature, and recognize that in a world of noisy megaphones, real attention is still a scarce thing.

If you want to go deeper, the authors have the details. They each bring a different stance to the same general puzzle. Read the ones that match the patch of code you care about, or the slice of business you find messy. Either way, you’ll find someone writing from a true place: practical frustration, nostalgic fondness, or clear-eyed strategy. And that’s worth reading.