The essay argues that every material downgrade — 420D nylon replacing 1000D Cordura, EVA swapped for molded foam, unbranded zippers for YKK, bar-tacks replacing stitched-through stress points — was a deliberate line item on a cost-reduction spreadsheet. Product managers with fixed margin targets and retail price points made conscious choices, knowing buyers shop by brand and photography rather than teardown analysis.
The essay's deeper thesis is that degradation happens when the people who notice quality aren't the people making purchase decisions. The backpack got worse not because nobody noticed, but because the noticers weren't the buyers — a structural diagnosis that applies wherever a product's buyers and users diverge.
The editorial extends Doctorow's platform-specific framework to physical products, arguing the backpack story is a field guide to enshittification written in the grammar of atoms rather than bits. The underlying dynamic — success metrics drifting away from what made the product good — is older and broader than platforms, and maps onto software with uncomfortable precision.
The top-voted comments in the 257-comment thread weren't about bags at all — they came from engineers who saw the same cost-cutting logic operating in software, hardware, and services they build. The pattern of deliberate quality regression driven by spreadsheet optimization resonated across industries far beyond consumer goods.
An essay titled *Your Backpack Got Worse on Purpose*, published on the newsletter Worse On Purpose, hit the front page of Hacker News this week with 289 points and a long comment thread. The piece catalogs, with receipts, how the backpack most people carry to work degraded over the last ten to fifteen years: thinner 420D nylon swapped in for 1000D Cordura, molded foam panels replaced with die-cut EVA, YKK zippers downgraded to unbranded equivalents, and stitched-through stress points replaced with bar-tacks that fail at the first full load.
The author's argument is not that manufacturers got lazy. It's that they got *deliberate*. Every substitution was a line item in a cost-reduction spreadsheet, signed off by a product manager who had a margin target, a retail price point that could not move, and a consumer base that buys primarily on brand and photograph rather than on a teardown. The backpack didn't degrade because nobody noticed. It degraded because the people who noticed weren't the people buying.
The HN thread, predictably, turned into a 400-comment referendum on everything from Patagonia's repair program to why the Jansport Superbreak from 2005 still works and the one your nephew got last Christmas is already fraying. But the comments that got the most upvotes weren't about bags. They were from engineers recognizing the pattern in their own industry.
The essay is really a field guide to enshittification written in the grammar of physical goods, and it maps onto software with uncomfortable precision. Cory Doctorow coined the term for platforms — the three-stage slide from "good to users" to "good to business customers" to "good to shareholders" — but the underlying dynamic is older and broader. It's what happens when a product's buyers and its users diverge, and when the people measuring success stop measuring the thing that made the product good in the first place.
Backpacks have that split: the buyer is a shopper in a store or a browser on a product page, making a five-minute decision based on colorway and brand. The user is the same person three years later, discovering that the laptop compartment's foam has compressed into a biscuit. The feedback loop between use and purchase is too slow and too diffuse to discipline the manufacturer. The manufacturer optimizes for the decision surface, not the use surface, because that's where the money is measured.
Sound familiar? Your IDE now ships with an AI sidebar you didn't ask for and can't fully remove. Your cloud provider's console added three new abstractions that exist to sell you a managed service you don't need. Your observability vendor raised prices 40% and called it a "platform unification." The database you picked three years ago for its boring reliability now has a shiny vector extension, a chat interface, and a roadmap slide that says "AI-native" on every bullet. None of this was incompetence either. It was all signed off by a PM with a margin target, a growth goal, and a dashboard that rewarded ARR over whether `CREATE INDEX` still finished in under a second.
The backpack essay is useful to developers not because backpacks matter but because physical degradation leaves evidence. You can weigh the new bag against the old. You can cut open the panel and photograph the foam. Software rot is harder to A/B because the binaries don't sit in a closet for a decade — but the same forces apply, and the same substitutions happen, just with less tactile proof. The dev-tool equivalent of thinner nylon is a free tier that quietly drops from 10 seats to 3, a CLI flag that becomes a paid feature, or a log retention window that shrinks from 30 days to 7 on "pro."
Community reaction on HN crystallized around one observation: the pattern is visible once you know to look for it, and it's accelerating in categories with high switching costs. Nobody re-platforms their auth provider casually. Nobody rewrites against a new time-series DB on a whim. The moment a vendor knows your migration cost exceeds their next price hike, the calculus inverts. The nylon gets thinner.
Treat every tool you depend on as a backpack that might be getting worse on purpose, and build the muscle to notice before you're locked in. Three concrete habits:
First, keep a dated snapshot of what the tool actually does. When you adopt a SaaS, write down the current quota, the current pricing tier, the current latency on your reference workload, the current feature set you rely on. Review that snapshot every six months. The drift is almost always negative and almost always gradual enough that you won't remember without the receipt. This is the software version of weighing the old bag.
Second, watch for the tell: when the buyer and the user split. In enterprise SaaS, the buyer is procurement or a VP; the user is you. The moment a vendor's sales motion is aimed upward of your desk, the product will start optimizing for demos, governance theater, and "AI-ready" slides, and away from the sharp, boring edges that made it good. If the changelog stops mentioning bug fixes and starts mentioning "strategic initiatives," the nylon is already being swapped.
Third, price in the exit. Open-source alternatives, self-hostable forks, and standards-based protocols aren't ideological preferences; they're options contracts against enshittification. You pay a premium in setup cost for the right to leave on your own timeline. When you're picking between a managed service and a well-maintained OSS equivalent, the managed service isn't just more expensive by the invoice — it's more expensive by the switching cost you're implicitly capitalizing.
The backpack essay is going to age well, because the dynamic it describes — cost-down in invisible places, feature-up in photogenic ones — is the default operating mode of any company past its founder-led phase. Software has been partially immune because competition was fierce and switching was cheap. Both of those conditions are weakening: consolidation is eating every dev-tool category, and every vendor is racing to add enough integration surface that leaving means a quarter of engineering time. Expect the nylon to keep getting thinner. Expect the zipper to keep getting cheaper. And expect the product page to keep insisting, with a straight face, that this year's model is the best one yet.
If you're looking for a backpack, I can't recommend Osprey enough. They are still a independent US company with a lifetime warranty they actually stand by. I had to call their customer service just last week after I ordered the wrong size bag. I was connected to an actual human immediately
Boots. I have bought Italian-made boots for 23 years - the same brand and model. 1. pair the inner lining died, rest was fine-to worn for 7 years, including laces, 2. pair lasted 7 years, the inner lining held this time, laces too. 3. pair was smaller than indicated, I stood the pain for one winter,
> I'll be writing about those next.I doubt it, you didn't write about this! You prompted it and signed your name to it.
Talk to any Aussie tech workers you know and chances are they or their friend has a Crumpler backpack. (No affiliation just a big fan).Originally created by a bike courier who sick of bags breaking sewed a bag out of the toughest material he could get, marine canvas. To this day they make somewhat i
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.
While I personally find this kind of thing extremely annoying, to me, the main problem is the _difficulty_ of determining quality. The Donut media guys did a (relatively unscientific) video comparing a whole bunch of products from the 50s to modern day across several price points. What they found wa