interpol_p

I think the DMA mandates that Apple not give Safari advantages over other browsers. Being able to run PWAs seems like it could be considered an advantage? Not sure though

They are not allowed to give their browser an advantage under the DMA. If you take a look at BrowserEngineKit and BrowserKit there is a significant API surface area they offer for third-party browser engines. They must have been building this for some time. It's really detailed, down to allowing developers to implement their own JIT! [1] they have custom UI components replacing their standard scroll views with ones that better support nested scrollable DOM elements. It's a staggering amount of engineering effort

I can totally believe that there is not enough time to re-think and re-architect how to implement push notifications, local storage and whatever other perks PWAs get for non-Safari third-party browser engines running as "apps." They may have lots of money and engineers, but throwing more of them at this problem is not going to build a well designed, thoroughly tested, and secure implementation any faster

[1]: https://developer.apple.com/documentation/browserenginekit/p...

It's pretty well hidden to the point of "why do they bother supporting this feature?"

I wonder if they use something like this for Apple Music

I notice a lot of the album art, playlist art, and more is animated. Much of it has abstract vector-style graphics. Seems like this would be a good place to start testing something like this

I have been using an iPhone 14 Pro. It's thick, I think I got used to it?

Recently someone handed me an old Samsung Galaxy S7 to use for testing. It was a revelation: I wanted and dearly missed having a phone that thin. I really hope someone pushes for thinness again. I miss the iPhone 6 (Edit: a big part of this might be down to the weight, the squared off edges, and the enormous camera bump)

Same thing with laptops. I updated from a MacBook Air M1 to a MacBook Pro 14". The 14" is fast enough to justify its bulk — but barely. Every time I pick up the Air it feels effortlessly portable, when I grab the MacBook Pro it just does not feel thin and light enough

As long as you _know the right answer_, ChatGPT is a very useful tool

90% of the time I know the code I want to write. ChatGPT just writes it for me, when I read it I can confirm it is what I would have written, or fix it if not (or more likely: tell ChatGPT to fix it)

This means I don’t have to google for syntax, canonical ways of doing things, etc.

If you are asking ChatGPT for facts, triple check everything it tells you. Or better still, paste the Wikipedia entry to it first, then ask for the style or type of writing you are interested in about those facts

Hey me too. I love Apple Arcade and when my kids want a game, my first question is, “did you look on arcade?”

However, my belief is the entire App Store should be subject to the same rules as Apple Arcade. It would be a far healthier app ecosystem, far less predatory

I don’t think it’s special for Spotify. It comes under the “Reader” apps clause (originally carved out for Kindle?) where apps which sell: music, movies, books, email can require that a user create an account on the web first (but cannot link to it, which is stupid)

Yeah I was just trying to make a cynical point that when they allow you to filter by no-in-app-purchases, they charge you for the privilege

Hah! Actually, they have kind of done this with Apple Arcade, and charge you a $10 subscription for the service

Just a note that for my business (and many others) Apple takes a 15% cut

Apple will lower their cut to 15% if you earn less than $1 million USD/year across your app businesses, or if you sell subscriptions and your subscribers persist for longer than 1 year (so 2nd - Nth year of subscriptions are split at 85/15)

Not defending the size of their commission, but in practice it does vary from the 30% that is commonly quoted

I don't think you can sign up for Spotify using in-app-purchase. Once you're in the app it says:

"You can't upgrade to premium in the app. We know, it's not ideal"

You have to go to their site to upgrade your account. Apple gets a 0% cut of Spotify's subscription revenue

Fair point. Though in my case it is also a business venture, but with very modest goals and building it consistently because I want it to exist

I guess my point is I just hope the author wasn't too quick to write off their attempt as a failure. Things can succeed after their first year, or even after their fifth

It really depends on your definition of "winning." My partner and I decided to make a small game together back in 2020, release it, and maintain it. "Winning" for me meant that it should earn $500 per year and have regular players who got the same joy from it that we got from making it. It was a really modest goal

We kept working on it over the last three years. It has just been slowly, but consistently getting better, with more players. Now it's earning $500/month

There's always the fantasy and the lure of building something that goes viral and becomes a sensation overnight, but there's something to be said for slow-and-steady, consistency, and doing something joyful!

Brother might be closest to this? See https://www.theverge.com/23642073/best-printer-2023-brother-...

I have a black-and-white laser Brother printer from over 15 years ago. Never changed the toner, comes out once every few months for the rare occasion I need to print something. I love it

Interesting, my kids all have 20+ large iMessage groups for their friends at school. They play Minecraft and Among Us while on FaceTime calls. They are in the 8 - 11 year range. So it is certainly down to demographics, but perhaps not age

It's not a brain. It's more like a very esoteric zip file which can extract a "lossy" version of the copyright material it was trained on, and reproduce that in an uncountable number of combinations that are meaningful to humans

In my family they are. I am in Australia and almost everyone I text has their phone number come up in blue, signifying iMessage/iPhone

For example, when RSVPing to a kid's birthday party, other parents' numbers are inevitably blue. When selling and buying items, the contacts for those sales have always been blue numbers, it's rare to encounter a number that doesn't "turn blue" when I enter it into the "to" field

I would say maybe 5% of the people I know and text use Android. For one of those people I use Signal, one other has asked me to use Facebook Messenger, one has asked me to use WhatsApp, and the remaining few use SMS. It's a pain to use three separate apps to message just these three people!

One of my cousins switched to an Android phone. This broke our long-standing group message in iMessage, so she was no longer able to be included in it. After two years of this her siblings simply ordered her a new iPhone and she is back in the group chat

Getting everyone to move their default messaging behaviour for one person is a huge ask. It was easier for one person to just relay the group chat info instead, but when this became annoying, it was even easier to buy her a new phone

From the article:

> How do we know it’s training data?

> How do we know this is actually recovering training data and not just making up text that looks plausible? Well one thing you can do is just search for it online using Google or something. But that would be slow. (And actually, in prior work, we did exactly this.) It’s also error prone and very rote.

>

> Instead, what we do is download a bunch of internet data (roughly 10 terabytes worth) and then build an efficient index on top of it using a suffix array (code here). And then we can intersect all the data we generate from ChatGPT with the data that already existed on the internet prior to ChatGPT’s creation. Any long sequence of text that matches our datasets is almost surely memorized.

Any significantly long sequence, repeated character-for-character is very unlikely to be generated and in there by pure coincidence. The samples they show are extremely long and specific

This is a really interesting exploration. The M3 Pro was a little confusing at first, having fewer performance cores than its predecessors, and more, but even slower clocked efficiency cores. But the power consumption graph clearly shows the benefits of this design decision, very cool chip

I use ChatGPT daily and have been coding for over two decades. I get it to write all the tedious stuff, and it basically writes all my internal tools, scripts and so on. The last big things it wrote for me were:

A macOS SwiftUI app to go through all the localization strings in my apps and show them in a table where I could see missing translations, and interactively translate them… by sending the, back to ChatGPT

Another macOS app to graph Google ngram data for word lists. It was great because I didn’t even need to check whether there was an API for ngrams, I just asked it for the function, then for the code to plot it in Swift Charts

Sometimes I consult it on refactoring problems (eg., this struct has new responsibilities, what would be a more appropriate name? How do you see it fitting into this section of the code?). It’s a great tool and easily worth the money

It almost never writes any “interesting” code, however

You’re totally right that there is now more complexity, especially around tooling and deployment. We have pipelines now, Docker containers, things that just aren’t simple and direct like a plain HTML+CSS website

However in my experience the best “front-end engineers” are designers. One of the best I’ve ever worked with called himself “Devsigner.” He was a passionate Android developer who knew the platform conventions better than the designers did, he was able to identify tiny flaws in designs (incorrect font weights, inconsistent padding, adherence to platform conventions), as well as larger flaws in the UX, for example, how to consistently and intuitively navigate between screens, when and where to place navigation elements, etc

Development wise, he was able to devise the app architecture clearly. Write end-to-end mock servers to perform UI testing, build the dev ops pipeline, manage app releases, and implement the UI using best practices

I am the same sort of designer/developer, and I expect that many people are capable of doing both things like this. They just need the opportunity and positive feedback to get there

Websites are more complex, but I expect people to be capable of learning complex things and to be aware of, and apply, design principles

I posted this not because I thought it would be inflammatory, but because I very much agreed with the point that design is not as far “to the left” as it should be

I started out as an animator/designer, and then became a developer. I became a developer because I wanted to write plugins for the 3D modelling software I used. Over the decades I have become extremely jaded at how companies I have worked for organise into “UX Design” and “Development” teams. It is utter nonsense

Everyone I have worked with, once trained, is capable of doing both — perhaps not everyone tackles design or development with the same passion, but teams who have comfort with both areas of app/web design and development often produce superior products

Now I am in a position to build a team at a company that is divided into “UX” and “Development.” I hope to ask people on the design team to install Xcode and Android Studio, to get comfortable making quick prototypes in Compose and SwiftUI, to update asset catalogues and colours, and to make branches and commit changes. Designers are genuinely excited about this

Similarly, I have started giving developers edit access to Figma, helping them produce design and visually communicate on a Figma board, tweak and suggest designs, and even build parts that can be approved by people on the “design team.” Developers have no problem doing this

I am hoping that the lines between “designer” and “developer” dissolve and we simply have coordinated teams, with talented people able to direct the vision for both aspects of a product

In my personal side projects everyone is treated as both designer and developer, no one “blindly implements” design (which I have experienced at many large companies) and no one creates design without understanding the nuances of the platform. This results in much less back-and-forth, and products that ship with more consistent UI design

You can have it both ways: the internal cameras can be used to compute attentiveness entirely within the car's local computer, any imagery could be immediately discarded after processing (or better, processed in a way where the imagery and algorithms are hardware-isolated from any computing modules connected to a network or permanent storage). This is similar to how Apple computes a successful face ID or fingerprint scan

Probably because your monitor’s OS is booting

I certainly wouldn’t call Apple a hardware company. Their software is as much a part of the story of their products as their hardware, if not more so in many areas.

It’s interesting that you make this point. Because Meta’s video is _all about_ the hardware: here’s new lens design, here’s high density screen.

Apple’s Vision Pro release has been all about the software: the interaction models, how to use SwiftUI to make immersive and windowed apps, the windowing model, accessibility, type readability, and every little detail related to UI

I’d argue Apple is equally software and hardware

You cannot download and execute native code on iOS, the sandbox prevents it. Memory protection ensures that pages marked as data cannot be executed

There is a special entitlement to allow some Apple apps and frameworks to do this — Apple’s JavaScriptCore can do JIT, which requires executing data pages. Similarly, Swift Playgrounds can compile to native code and execute that code

So, for example, you could ship a C compiler on iOS and it would compile perfectly valid binaries, but you would never be able to execute those binaries natively without code-signing them and deploying them in an Apple sanctioned way, which you cannot do on-device (except with the above mentioned exceptions)

It’s really dependent on what reviewer you get. I develop a coding app for iPad (Codea) and for the longest time was not allowed to let users share their code in any form. I asked the reviewer “but they can just copy and paste it?” And the reviewer said “that is ok.”

In the end, it was down to whether it _looked like_ your app was downloading executable code more than whether it actually did. Reviewers are non-technical, and when the policy is technical they often make an inference that may be incorrect — however they will not budge on that without consulting actual engineers further up the chain, which can take weeks

They have become a lot more relaxed about this policy in the last five years or so (ever since Shortcuts)

I’ve been using Stage Manager on macOS for this. I have groups of apps, each one in a separate “stage” that I switch between to work on different things. Out of all the desktop paradigms I’ve tried, this one is working the best for me

I can constantly see the thumbnail groupings of apps that I can switch to along the left hand side of the screen. I’ve used other modes (like multiple desktops) but would often forget what spaces I had created, and ended up re-creating them without realising