1,248 karma · joined August 17, 2025
Sounds good.
I use iStat and I don’t think a week goes by without being alerted to some process keeping a CPU core busy for half an hour. It would be reasonable for something sensible such as photo tagging or disk indexing. But the many background processes can get stuck, just look online for the flood of complaints.
corespotlightd, for example, uses >80% CPU (of a core, I think) for long periods of time. I’m talking _days_. The multiple store and iCloud agents also seem to get stuck. If you inspect a process, you can sometimes see it’s been doing something intensive for days or even weeks.
I get notified at 10 minutes of intense usage, but then I just keep looking and it’s really, really typical for a process to spend a few hours doing something, nuking the battery life of my laptop. I’ve been on macOS since 2008 or so and it’s not getting any better when it comes to zombie processed.
I’ve reported issues to the DPA (Data Protection Agency) of Belgium and they handled it reasonably well, which I didn’t expect either.
Marketing tactics are part of consumer issues and many countries seem to have good consumer protection laws. Of course Grammarly’s issue was towards/against a business, and that colours the problem. I think the damage to their reputation is enough punishment, but there might be other ways to tackle this.
I merely fail to see the point of “oh well, the Humanity, what can we do”. Apathy is death.
Cynicism about public institutions is overrated.
I don’t see the need to have multiple trail shoes on offer, some of which have a non-replaceable sole. Lone Peaks are almost perfect, keep the stone guard, drop the total stack, and we’re cooking. The longevity of the shoes seems to be compromised by the fancy fabric in some areas, though.
Good luck with the Olympus. Don’t let the lack of prosumer foam discourage you!
My Altras last about 2x any other running shoe, and I even took them on trail runs (Lone Peaks).
Not defending the company, they’re pricey and the catalog is exaggerated, full of nonsense 30mm+ stack shoes, etc.
I’ve worked places where UX Research was telling us exactly what people ask for, but because it has to pass through multiple layers of PM, the end result isn’t what people asked for.
E.g. “families are telling us they want to talk directly with caregivers on WhatsApp” -> we build an admin dashboard with live chat. Nobody used the thing.
Perhaps there is a relationship between layers of abstraction at a business and clarity of product. The more humans in the chain — with their own biases and agendas —, the less likely you are to build what people want.
“People don’t know what they want” is a bit lame. Perhaps they just don’t want your ad-fueled growth-marketed product.
I get it, but at 9–10% error rate, you’re basically gambling with the results.
Source: https://www.nytimes.com/2026/04/07/technology/google-ai-over...
If I enjoy an album, I’ll listen to it some 15-20 times before picking favourite songs or simply moving on. What you’re describing sounds more like consuming albums, not enjoying them. My 2 cents :)
Also: if you’re going to charge laptop-grade prices for a phone because it… gets bigger? Well, better make it a really good experience, i.e. no crease that develops fractures later.
I agree that SQLite might not be the better choice given many Jellyfin users are probably on ZFS or BTRFS.
(…but you couldn’t call me capitalistic, that really doesn’t suit me.)
I just hope they can continue to develop the product, because I’m using it daily and don’t want it to go away.