I’m in a camp where Tahoe is just ugly, but worked fine on my mac and with my usage and devices. On the other hand, iOS 26 is so incredibly bad. Anything from anti-user choices, UI glitches, to keyboard… it’s tiring. It might be the one where I’m pushed to a different OS for the first time. I doubt Apple can fix itself in the near future, with the leadership they have.
I run all the betas on my macbook air just to see where things are going. When I first installed Tahoe, I thought to myself, I guess it's just the beta, there's no way they are going to keep this look...
My workstation is pinned firmly to sequoia right now and will be until it's untenable or they undo their new UI design direction, which is, as you pointed out, ghastly.
Good read. If I imagine all these layers of Agent to Agent economic automations, I’m hopeful for a new class of technopriest hackers that are able to exploit nuances and weak security. And build new hidden empires within those systems.
I’d like to hear someone from MS if this has any impact, or if the report is overblowing it? Closing a library and pushing something called “Skilling” Hub sounds cartoonishly stupid and disrespectful.
It’s very close on their quality IMHO. Miles better on the value. I bought a Lumibricks set this year, after decades of buying Lego.
The set design (Cafe) was nice and it makes a good display piece. I liked their build techniques and integrated lights. The individual bricks were clean and with good color uniformity. But after a life of building Lego sets, I was able to quickly tell that tactile forces for connecting pieces were different to Lego. I will buy more of their sets, but I will keep buying Lego too.
Sorta, but what I have in mind is a graphical browser, not exactly like the OP. It would have CSS support under the user's control, not the website's control. Navigation would also be more constrained and predictable than the totally open-ended nature it has today. That would enable things like perfect universal dark mode, and excellent accessibility. Fuzzy navigation is on my wishlist too, based on a predictable nav structure.
Waratek https://waratek.com | Principal Java Engineer | Hybrid in Dublin or Remote in Europe
We are working on a specialised Cybersecurity Runtime Application Self-Protection (RASP) tooling.
We are seeking an exceptional programmer with expertise in Java, someone who will thrive in exploring the JDK ecosystem further. You will join a team developing the core technology behind our Java Agents, working on cutting-edge solutions to enhance security.
Feel free to reach out with questions or applications! My email is in my HN profile.
Shellspec is delightful to use. Every time I go to test with it I end up finding bugs. But also every time I go to test with it I end up convincing myself to not use bash. I imagine the same would be true of bash_unit.
This was exactly the reason I’ve abandoned them. As mentioned in other comments, they look like a good way to create a design/component system in them, because of the interop with any frontend JS framework (who doesn’t want that?!). So you deal with clunky CSS sharing and theming capabilities and there might be some solutions on the horizon. But now you want to build a static website with it (landing page, docs, blog…) and you can’t pre-render them, because there is no declarative Shadow DOM element https://github.com/w3c/webcomponents/blob/gh-pages/proposals... (rejected https://github.com/whatwg/dom/issues/510). Yeah, you can pre-render them without the Shadow DOM. Then if you want to make those chunks of HTML into WCT on load (sort of rehydration), it will trigger repaints, as you have to remove the chunk of HTML, put it into WCT and parse everything again. Or you don’t put it into Shadow Root, to prevent repaint - but at that point, if your components can work this way, why even use WCT?
reply