Luna Display is adding Windows, which, yeah, of course they are. That was exactly the right move to make. Apple sherlocked their main feature — treating an iPad as an additional display for your Mac — as "Sidecar" in macOS 10.15 Catalina, and they dit the thing that makes sense for everyone: cover the 90% use case for their own user base. This leaves Astropad/Luna Display to figure out how to focus on the remaining 10%, and they have.
(via Miguel, who calls it "turning a checkbox feature into a full product offering")
Apparently old, but a good example of discoverability: when Google Translate's UI was changed from unlabeled buttons that show a camera, a microphone, and a weird squiggle to four labeled buttons for Camera, Handwriting, Conversation, and Voice, usage supposedly went up 25%.
I'd even say the new UI looks more attractive, but I'm guessing minimalists would disagree.
Now that iOS has a mechanism where you can choose whether to upgrade to 15 or stick to 14 for a while, I wonder if one of the next steps, two or three years down the road, will be to allow developers to dual-boot.
It would increase complexity for Apple, and the attack vector, too, but it would also help the developer story. Rather than either having to buy a separate device to run betas on (or older OSes you still want to support), you use the same device and boot into it.
iOS 15's Live Text keeps being impressive. (Or concerning, for some.)
The democratization of OCR, if you will.
Here we have student A taking notes on a laptop, and student B, a few rows back, using an iPhone to take a phot of A's laptop. iOS 15 then automatically OCRs the photo, so B can highlight and copy the text. Just like that.
It's not the technology that's surprising, it's the seamless implementation.
Another take: 10 years from now, kids might not even get how this is impressive. Of course computers would be able let you copy text from photos; why not?
An evil pitfall in recent C# versions:
public Guid Foo => Guid.NewGuid();
vs.
public Guid Foo { get; } = Guid.NewGuid();
Do they differ? Yes.
The first expands to:
public Guid Foo { get { return Guid.NewGuid(); } }
And once you read that, you might notice the problem: a new GUID isn't generated when the class is instantiated, but rather each time the property is accessed.
Microsoft showed off a refreshingly different take on emoji in July, but at least for now, that's not quite what they'll ship.
Most of the gradients and textures seem gone in favor of a much flatter, duller look.
My guess is either the team commissioned to design these never communicated with the one implementing them in Windows, or they were optimistic they could implement these, but have run into delays or insurmountable issues. Currently, emoji on Windows work as multiple layers of colored glyphs, which might make more complex 3D shapes (or even just linear gradients) hard to implement.
Perhaps there were plans to go with a different approach, and those had to be ditched at the last minute.
It's also possible they tested poorly in accessibility.
Other platforms have different approaches. For example, Apple and Google simply embed PNG graphics into a font; this comes with the downside that they're not vector graphics at all, but in practice makes for higher graphical fidelity. Adobe and Mozilla have created an SVG-based approach, which in turn might come with performance/efficiency concerns. Because eveyone rushed to market way back then, none of these formats are compatible, and at least four exist.
(via)
Speaking of emoji, while I mostly think Apple's are fine, the MSN emoticons for hugging back in the day are way better than what Apple has to offer.
Thingiverse, a major website for designs you can 3D-print, got breached.
(via)
Spelling or pronouncing Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch is… fun.
(via)
Last-minute rumors are appearing that we'll see a "notch" on next week's MacBook Pros.
(via)