Their primary business goal is to sell hardware. Yes, they’ve diversified into services and being a shopping mall for all, but it is about selling luxury hardware.

The promise of privacy is one way in which they position themselves, but I would not bet the bank on that being true forever.

The alternative is Google / Android devices and OpenAI wrapper apps, both of which usually offer a half baked UI, poor privacy practices, and a completely broken UX when the internet connection isn't perfect.

Pair this with the completely subpar Android apps, Google dropping support for an app about once a month, and suddenly I'm okay with the lesser of two evils.

I know they aren't running a charity, I even hypothesized that Apple just can't build good services so they pivoted to focusing on this fake "privacy" angle. In the end, iPhones are likely going to be better for edge AI than whatever is out there, so I'm looking forward to this.

No, the alternative is Android devices with everything except firmware built from source and signed by myself. And at the same time, being secure, too.

You just can't have this on Apple devices. On Android side choices are limited too, I don't like Google and especially their disastrous hardware design, but their Pixel line is the most approachable one able to do all these.

Heck, you can't even build your own app for your own iPhone without buying another hardware (a Mac, this is not a software issue, this is a legal issue, iOS SDK is licensed to you on the condition of using on Apple hardware only) and a yearly subscription. How is this acceptable at all?

These features work on all the modern desktop browsers and on Android tho!

WebGPU isn't standardized yet. Hell, most of the features people complain about aren't part of any standard, but for some reason there's this sense that if it's in Chrome, it's standard - as if Google dictates standards.

Realistically, given the market share of Chrome and Chromium based browsers, they kind of do.

Because they set the terms of use of the SDK? You're not required to use it. You aren't required to develop for iOS. Just because Google gives it all away for free doesn't mean Apple has to.

Do you have a legal right to write software or run your own software for hardware you bought?

Because it’s very easy to take away a right by erecting aritificial barriers, just like how you could discriminate by race at work, but pretend you are doing something else,

I've never heard of such a thing. Ideally I'd like that, but I don't have such freedoms with the computers in my cars, for example, or the one that operates my furnace, or even for certain parts of my PC.

You aren't allowed to repair the "thing' because the software can detect you changed something and will refuse to boot. And whenever it suits the manufacturer, they will decide when the 'thing' is declared out of support and stops functioning.

I would say you are not an owner then, you (and me) and just suckers that are paying for the party. Maybe it's a lease. But then we also pay when it breaks, so it more of a digital feudalism.

No, obviously not. Do you have a right to run a custom OS on your PS5? Do you have a right to run a custom application on your cable set-top box? Etc. Such a right obviously doesn’t exist and most people generally are somewhere between “don’t care” and actively rejecting it for various reasons (hacking in games, content DRM, etc).

It’s fine if you think there should be, but it continues this weird trend of using apple as a foil for complaining about random other issues that other vendors tend to be just as bad or oftentimes even worse about, simply because they’re a large company with a large group of anti-fans/haters who will readily nod along.

Remember when the complaint was that the pelican case of factory OEM tools you could rent (or buy) to install your factory replacement screen was too big and bulky, meaning it was really just a plot to sabotage right to repair?

https://www.theverge.com/2022/5/21/23079058/apple-self-servi...

Yes, I do. That was and continues to be a valid complaint, among all other anti-repair schemes Apple have come up with over the years. DRM for parts, complete unavailability of some commonly repaired parts, deliberate kneecapping of "Apple authorized service providers", leveraging the US customs to seize shipments of legitimate and/or unlabeled replacement parts as "counterfeits", gaslighting by official representatives on Apple's own forums about data recovery, sabotaging right to repair laws, and even denial of design issues[1] to weasel out of warranty repair just to name a few.

All with the simple anti-competitive goal of making third party repair (both authorized and independent) a less attractive option due to artificially increased prices, timelines to repair, or scaremongering about privacy.

https://arstechnica.com/gadgets/2022/12/weakened-right-to-re...

https://www.pcgamer.com/ifixit-says-apples-iphone-14-is-lite...

[1] Butterfly keyboards, display cables that were too short and failed over time

No, it doesn’t - because you can simply not use the tools if you don’t want. You can just order a $2 spudger off Amazon if you want, you don’t need the tools at all.

It continues to be a completely invalid complaint that shows just how bad-faith the discussion about apple has become - it literally costs you nothing to not use the tools if you want, there is no downside to having apple make them available to people, and yet you guys still find a way to bitch about it.

Moreover, despite some “bold” proclamations from the haters… no android vendors ever ended up making their oem tooling available to consumers at all. You have to use the Amazon spudger on your pixel, and you will fuck up the waterproofing when you do your repair, because the android phone won’t seal properly against water without the tools either. IPX sixtywho!?

It’s literally a complete and total net positive: nothing was taken away from you, and you don’t need to use it, and it makes your life easier and produces a better repair if you want it. Apple went out of their way to both make the tooling available to normies who want to rent it or people who want to buy it for real. And people still bitch, and still think they come off better for having done so. Classic “hater” moment, in the Paul Graham sense. Anti-fanboys are real.

https://paulgraham.com/fh.html

Literally, for some people - the pelican cases with the tools are too big and heavy. And that’s enough to justify the hate.

Again, great example of the point I was making in the original comment: people inserting their random hobby horse issues using apple as a foil. You don’t like how phones are made in general, so you’re using apple as a whipping boy for the issue even if it’s not really caused or worsened by the event in question etc. Even if the event in question is apple making that issue somewhat better, and is done worse by all the other vendors etc. Can’t buy tooling for a pixel at all, doing those repairs will simply break waterproofing without it, and you’re strictly better off having the ability to get access to the tooling if you decide you want it, but apple offering it is a flashpoint you can exploit for rhetorical advantage.

I think the dirty little secret here is that an iPhone is just about the only phone, apart from maybe some of the really nice Google and Samsung flagships, that anyone wants to repair, because they're bloody expensive. Which is fine and dandy but then do kindly park your endless bemoaning of the subjects of e-waste and non-repairable goods, when Android by far and away is the worse side of that equation, with absolute shit tons of low yield, crap hardware made, sold, and thrown away when the first software update renders it completely unusable (if it wasn't already, from the factory).

Perhaps you haven't noticed, but once you tally up overpriced parts together with their oversized, heavy, expensive rental of tools that you don't need, you end up with a sum that matches what you would pay to have it repaired by Apple - except you're doing all of the work yourself.

A curious consumer who has never repaired a device, but might have been interested in doing so, will therefore conclude that repairing their own device is 1. Far too complicated, thanks to an intimidating-looking piece of kit that they recommend, but is completely unnecessary, and 2. Far too expensive, because Apple prices these such that the repair is made economically nonviable.

So yes, I still believe that this is Apple fighting the anti-repair war on a psychological front. You're giving them benefit of the doubt even though they've established a clear pattern of behavior that demonstrates their anti-repair stance beyond any reasonable doubt - although you dance around the citations and claim that I'm being unreasonable about Apple genuinely making the repair situation "better".

Futhermore, I'm not a fanboy or anti-fanboy of any company. The only thing I'm an anti-fanboy of are anti-consumer practices. If Apple changed some of their practices I'd go out and buy an iPhone and a Macbook tomorrow.

The fact that I pointed out that Apple is hostile against repair does not mean that I endorse Google, Samsung, or any other brand - they all suck when it comes to repair, yet you're taking it as a personal attack and calling me names for it.

It can be difficult to handle the intimation that maybe there is the mirror image of the "brainless apple sheeple" too. Haters exist - people who are not able to approach a situation fairly and are just relentlessly negative because they hate X thing too. Negative parasocial attachment is just as much of a thing as a positive one.

And when you get to the point of "literally making the tools available is bad because the pelican cases are too heavy" you have crossed that rubicon. I am being very very polite here, but you are not being rational, and you are kinda having an emotional spasm over someone disagreeing with you on the internet.

Yes, if you want OEM parts and you rent OEM tooling it's probably going to come close to OEM cost. That isn't discriminatory, if the prices are fair+reasonable, and objectively they pretty much are. $49 to rent the tools, and have them shipped both ways, etc, is not an unreasonable ask.

Not having a viable business model for your startup doesn't mean the world is wrong. It means you don't have a viable business idea. And yeah, if you are renting the tools as a one-off, and counting your personal time as having some value (or labor cost in a business), then you probably are not going to get costs that are economical with a large-scale operator with a chain operation and an assembly-line repair shop with repair people who do nothing but repair that one brand. That's not Apple's fault.

What we ultimately come down to with your argument is "apple is killing right-to-repair by being too good at repair and providing too cheap repairs such that indie shops can no longer make a margin" and I'm not sure that's actionable in a social sense of preventing e-waste. People getting their hardware repaired cheaply is good. Long parts lifetimes are good. Etc.

Being able to swap in shitty amazon knockoff parts is a whole separate discussion, of course. And afaik that is going to be forced by the EU anyway, consequences be damned. So what are you complaining about here?

https://news.ycombinator.com/newsguidelines.html

https://news.ycombinator.com/newsguidelines.html

edit: Describing others as "bitching", "bad faith", and "hater", runs afoul of half of the guidelines on this site. That comment somehow isn't moderated, but mine is called out for crossing into flamewar?

Even if I hadn't, though, we need you to follow the rules regardless of what anybody else is doing, and the same goes for every user here. Pointing a finger at the other person isn't helpful.

As noxious as that whole thing is, it’s literally the law. I agree the outcome is horrifying of course… stallman was right all along, it’s either your device or it’s not.

And legally speaking, we have decided it’s ok to go with “not”.

Sure, as a SWE I'm not going to buy a computer unable to run my own code. A smartphone is an ergonomic portable computer, so I say no to iPhone and would like to remind others who didn't have a deep think into this about it.

Because as you described, the only alternatives that exist are terrible experiences for basically everyone, so people are happy to pay to license a solution that solves their problems with minimal fuss.

Any number of people could respond to “use Android devices with everything except firmware built from source and signed by myself” with the same question.

Normal users will not do this. Just because many of the people here can build and sign a custom Android build doesn't mean that is a viable commercial alternative. It is great that is an option for those of us who can do it, but don't present it as a viable alternative to the iOS/Google ecosystems. The fraction of people who can and will be willing to do this is really small. And even if you can do it, how many people will want to maintain their custom built OSes?

Unfortunately a lot of the "freedom" crowd think that unless you want to be an 80s sysadmin you don't deserve security or privacy. Or computers.

a maxim that seems to hold true at every level of computing is that users will not care about security unless forced into caring

with privacy they may care more, but they are easily conditioned to assume it's there or that nothing can be realistically be done about losing it

There are a couple caveats:

1. It is still a bit tricky for a non-technical person to install. Should not be a problem if they know somebody who can help, though. There's been some progress making the process more user friendly recently (e.g. WebUSB-based GrapheneOS installer).

2. There are some papercuts if you don't install Google services on your phone. microG [2] helps with most but some still remain. My main concern with this setup is that I can't use Google Pay this way, but having to bring my card with me every time seems like an acceptable trade off to me.

[1]: https://divestos.org/

[2]: https://microg.org/

I wouldn't bet on this long term, since it fully relies on Google hardware, and Google's long-term strategy is to remove your freedom piece by piece, cash on it, not to support it.

The real alternative is GNU/Linux phones, Librem 5 and Pinephone, without any ties to greedy, anti-freedom corporations.

There are people who don't know how to use file explorer, new generation grows up in a world of iPhones without ever seeing file system. Any other bright ideas?

Apple's definitely pulling a "Heinz" move with privacy, and it is true that they're doing a better job of it overall, but Google's not completely horrible either.

https://store.google.com/intl/en/ideas/articles/pixel-featur...

In any case iPhone probably don't have enough memory to run a 3.25B model? e.g. 15 pro only have 8 GB (and Gemini Nano seems to only work on the 12GB Pixel 8 Pro) and 14 has only 6GB, that hardly seems sufficient for even a small LLM if you still you want to run the full OS and other apps at the same time.

What exactly are you expecting? The current hype for AI is large language models. The word 'large' has a certain meaning in that context. Much larger that can fit on your phone. Everyone is going crazy about edge AI, what am I missing?

Distributed mixture of experts sounds like an idea. Is anyone doing that?

I’m imagining a lot of these LLM products on phones will be used for live translation. Imagine a large crowd event of folks utilizing live AI translation services being told completely false translations because an actor deployed a 51% attack.

No point asking GPT4 “what time does John’s party start?”, but a local LLM can do better.

If you clone a model and then bake in a more expensive model's correct/appropriate responses to your queries, you now have the functionality of the expensive model in your clone. For your specific use case.

The size of the resulting case-specific models are small enough to run on all kinds of hardware, so everyone's seeing how much work can be done on their laptop right now. One incentive for doing so is that your approaches to problems are constrained by the cost and security of the Q&A roundtrip.

Care to cite these subpar Android apps? The app store is filled to the brim with subpar and garbage apps.

>Google dropping support for an app about once a month

I mean if you're going to lie why not go bigger

>I'm okay with the lesser of two evils.

So the more evil company is the one that pulled out of China because they refused to hand over their users data to the Chinese government on a fiber optic silver plate?

They also had project DragonFly if you remember.

The lesser of two evils is that one company doesn’t try to actively profile me (in order for their ads business to be better) with every piece of data it can find and forces me to share all possible data with them.

Google is famously known to kill apps that are good and used by customers: https://killedbygoogle.com/

As for the subpar apps: there is a massive difference between the network traffic when on the Home Screen between iOS and Android.

The Chinese government has access to the iCloud account of every Chinese Apple user.

>They also had project DragonFly if you remember.

Which never materialized.

>The lesser of two evils is that one company doesn’t try to actively profile me (in order for their ads business to be better) with every piece of data it can find and forces me to share all possible data with them.

Apple does targeted and non targeted advertising as well. Additionally, your carrier has likely sold all of the data they have on you. Apple was also sued for selling user data to ad networks. Odd for a Privacy First company to engage in things like that.

>Google is famously known to kill apps that are good and used by customers: https://killedbygoogle.com/

Google has been around for 26 years I believe. According to that link 60 apps were killed in that timeframe. According to your statement that Google kills an app a month that would leave you 252 apps short. Furthermore, the numbers would indicate that Google has killed 2.3 apps per year or .192 apps per month.

>As for the subpar apps: there is a massive difference between the network traffic when on the Home Screen between iOS and Android.

Not sure how that has anything to do with app quality, but if network traffic is your concern there's probably a lot more an Android user can do than an iOS user to control or eliminate the traffic.

Most of the "Services" on that list are effectively apps, too:

VPN by Google One, Album Archive, Hangouts, all the way back to Answers, Writely, and Deskbar.

I didn't touch hardware, because I think that should be considered separately.

The first of 211 services on that site was killed in 2006.

The first of the 60 apps on that site was killed in 2012.

So even apps alone, 4.28 a year.

But more inclusively, 271 apps or services in 17 years is ~16/year, over one a month.

You need to remind yourself of the site guidelines about assuming the worst. Your comments just come across condescendingly.

Even with the additional apps you've selected it still doesn't come close to the one app per month claim.

>I didn't touch hardware, because I think that should be considered separately.

So why even mention it? Is Apple impervious to discontinuing hardware?

>The first of 211 services on that site was killed in 2006.

So we're talking about services now? Or apps? Or apps and services? The goal posts keep moving.

>You need to remind yourself of the site guidelines about assuming the worst. Your comments just come across condescendingly.

I suggest you also consult the guidelines in regards to calling people names. My comments were never intended to be inferred that way.

Source?

Google podcasts, Stadia, shitton of other discontinued applications?

Providing services is arguably at odds with privacy since a service with access to all the data can provide a better service than one without so there will be a tension between trying to provide the best services, fueling their growth, and privacy.

My point was that it's interesting how we can frame a service business "extracting fees" to imply wrongdoing. When it's pretty normal for all services to charge ongoing fees for ongoing delivery.

And as someone who presumably provides services for a living, what additional regulations would you like to be subject to?

Seems even more unlikely for non technical users.

It's just their latest market campaign, as far as I can tell. The vast majority of people buy iPhones because of the status it gives.

Marketing aside, all indications point to the iOS platform being the most secure mobile option (imo).

Theres no “status” to a brand of phone when the cheapest point of entry is comparable and the flagship is cheaper than the alternative flagship.

Marketing in most of europe is chiefly not the same as the US though so maybe its a perspective thing.

I just find it hard to really argue “status” when the last 4 iPhone generations are largely the same and cheaper than the Samsung flagships.

At Elgiganten a Samsung S24 Ultra is 19,490 SEK[0].

The most expensive iPhone 15 pro max is 18,784 SEK at the same store[1].

[0]: https://nya.elgiganten.se/product/mobiler-tablets-smartklock...

[1]: https://nya.elgiganten.se/product/mobiler-tablets-smartklock...

Rich people ask for the latest most expensive iPhone even if they're only going to use WhatsApp and Instagram on it. It's not because of privacy or functionality, it's simply to show off to everyone they can purchase it. Also to not stand out within their peers as the only one without it.

As another content said: it's not an argument, it's a fact here.

There's obviously tons of people who see iPhone as a status item. They're right, because iPhone is expensive and only the rich can buy them. This doesn't mean iPhone is not the best option out there for a person who doesn't want to extensively customize his phone and just use it.

It’s not about if you’ve used android, it’s about if you’ve beeen poor-ish or stingy

To some people those are luxuries- the most expensive phone they buy is a mid-range Motorola for $300 with snapdragon 750g or whatever. They run all the same apps after all, they take photos.

iPhones are simply outside of your budget.

Sure, some people pick up the iPhone because they like the specs, or the apps, or whatever else. That's why I said the majority picks it up for status, not all. But keep in mind nobody's judging the iPhone's specs or capabilities here. We're talking about why people buy it.

Ask any teenager why they want an iPhone. I'd be very surprised if even one said it's because of privacy. It's because of the stupid blue bubble, which is a proxy for status.

I'm pretty sure if Apple released the same phone again with a new name and design, people would still buy it. For the majority, it's not because of features, ease of use, specs, etc: it's status.

It’s like how people who don’t like US and want nothing to do with US still discuss US politics, because it has so much effect everywhere.

(Ironically no enough people discuss China in any coherent level of understanding)

Apple doesn't release new products, they wait until the actual brave and innovating companies have done the exploration and then capitalize on all of their learnings. Because they are never the first movers and they have mountains of cash, they're able to enter the market without the baggage of early adopters. They don't have to worry about maintaining their early prototypes.

Apple doesn't innovate or show leadership, they wait until the innovators have proven that the market is big enough to handle Apple, then they swoop in with a product that combines the visions of the companies that were competing.

Apple is great at what they do, don't get me wrong. And swooping in when the market is right is just good business. Just don't mistake that for innovation or leadership.

Granted, you may have been leaning very heavily on the dictionary definition of "comparable", in that the two numbers are able to be compared. However, when the conclusion of that comparison is "More than twice the price", I think you should lead with that.

Keep in mind, the iPhone SE is using a 3 year old processor, the Samsung A15 was released 5 months ago with a brand new processor.

According to various sites, the Mediatek Dimensity 6100+ is a 6nm update to a core that was released 3 years ago (Dimensity 700 on a 7nm). It's 5-10% faster, likely due to the update from 7 to 6nm, as the cores are the same and run at the same speed. It contains an updated bluetooth chipset (from 5.1 to 5.2) and supports a larger max camera. The camera on the A15 is well below the max size of the previous chipset, however, the increased camera bandwidth should ensure that the camera feels snappier (a common complaint on low-end phones). The process improvement should increase efficiency as well, however, there are not benchmarks that are able to test this.

But not only that, Apple CPUs are objectively leagues ahead of their competition in the mobile space. I am still using a IPhone released in 2020 with absolutely no appreciable slow down or losses in perceived performance. Because even a 4 years old IPhone still has specs that don't lag behind by much the equivalent Android phones, I still receive the latest OS updates, and because frankly, Android OS is mess.

If I cared about status, I would have changed my phone already for a new one.

My Pixel 4a here is also going strong, only the battery is slowly getting worse. I mean, it's 2024, do phones really still get slow? The 4a is now past android updates, but that was promised after 3 years. But at 350 bucks, it was like 40% less than the cheapest iPhone mini at that time.

Hardware is pretty beefed up but bloat keeps on growing, that is slowing things down considerably.

This is a lie. The latest Android SoCs are just as powerful as the A series.

>Because even a 4 years old IPhone still has specs that don't lag behind by much the equivalent Android phones, I still receive the latest OS updates, and because frankly, Android OS is mess.

Samsung and Google offer 7 years of OS and security updates. I believe that beats the Apple policy.

On the second part:

https://en.wikipedia.org/wiki/IPadOS_version_history

The last iPads to stop getting OS updates (including security, to be consistent with what Samsung and Google are pledging) got 7 and 9 years of updates each (5th gen iPad and 1st gen iPad Pro). The last iPhones to lose support got about 7 years each (iPhone 8 and X). 6S, SE (1st), and 7 got 9 and 8 years of OS support with security updates. The 5S (released in 2013) last got a security update in early 2023, so also about 9 years, the 6 (2014) ended at the same time so let's call it 8 years. The 4S, 2011, got 8 years of OS support. 5 and 5C got 7 and 6 years of support (5C was 5 in a new case, so was always going to get a year less in support).

Apple has not, that I've seen at least, ever established a long term support policy on iPhones and iPads, but the numbers show they're doing at least as well as what Samsung and Google are promising to do, but have not yet done. And they've been doing this for more than a decade now.

EDIT:

Reworked the iOS numbers a bit, down to the month (I was looking at years above and rounding, so this is more accurate). iOS support time by device for devices that cannot use the current iOS 17 (so the XS and above are not counted here) in months:

  1st - 32
  3G  - 37
  3GS - 56
  4   - 48
  4S  - 93
  5   - 81
  5C  - 69
  5S  - 112
  6   - 100
  6S  - 102
  SE  - 96
  7   - 90
  8   - 78
  X   - 76

The 8 and X look like regressions, but their last updates were just 2 months ago (March 21, 2024) so still a good chance their support period will increase and exceed the 7 year mark like every model since the 5S. We'll have to see if they get any more updates in November 2024 or later to see if they can hit the 7 year mark.

These are very disingenuous numbers that don't tell the complete story. An iPhone 7 getting a single critical security patch does not take into account the hundreds of security patches it did not receive when it stopped receiving support. It received that special update because Apple likely was told or discovered it was being exploited in the wild.

Google and Samsung now offer 7 years of OS upgrades and 84 months of full security patches. Selectively patching a phone that is out of the support window with a single security patch does not automatically increase its EOL support date.

Neither of these hopes retroactively invalidates the fact that Apple has had a much better track record of supporting old phone models up to this point. Even if you do split hairs about the level of patching some models got in their later years, they still got full iOS updates for years longer than most Android phones got any patches at all, regardless of severity.

This is not an argument that somehow puts Android on top, at best it adds nuance to just how much better iOS support has been up to this point.

Let's also not forget that if Apple wasn't putting this kind of pressure on Google, they wouldn't have even made the promise to begin with, because it's clear how long they actually care to support products with no outside pressure.

What we can do today is note that Apple never made a promise, but did provide very long security support for their devices despite that. They've already met or come close to the Samsung/Google pledge (for one device) on almost half their devices, and those are all the recent ones (so it's not a downward trend of good support then bad support, but rather mediocre/bad support to improving and increasingly good support).

Another fun one:

iPhone XS was released in September 2018, it is on the current iOS 17 release. In the absolute worst case of it losing iOS 18 support in September, it will have received 6 full years of support in both security and OS updates. It'll still hit 7 years (comfortably) of security updates. If it does get iOS 18 support in September, then Apple will hit the Samsung/Google pledge 5 years before Samsung/Google can even demonstrate their ability to follow through (Samsung has a chance, but Google has no history of commitment).

I have time to kill before training for a century ride:

Let's ignore everything before iPhone 4S, they had short support periods that's just a fact and hardly worth investigating. This is an analysis of devices released in 2011 and later, when the phones had, mostly, matured as a device so we should be expecting longer support periods. These are the support periods when the phones were able to run the still-current iOS versions, not counting later security updates or minor updates but after the major iOS version had been deprecated. As an example, for the iPhone 4S it had support from 2011-2016. In 2016 its OS, iOS 9, was replaced by iOS 10. Here are the numbers:

  4S       - 5 years
  5        - 5 years
  5C       - 4 years (decreased, 5 hardware but released a year later in a different case)
  5S       - 6 years
  6        - 5 years (decreased, not sure why)
  6S       - 7 years (hey, Apple did it! 2015 release, lost iOS upgrades in 2022)
  SE(1st)  - 5 years (like 5C, 6S hardware but released later)
  7        - 6 years (decreased over 6S, not sure why)
  8        - 6 years
  X        - 6 years

BTW, my daily driver is an iPhone 13 and I was coming from an iPhone X. So I'm well aware of the incredible support Apple provides its phones. Although, I would still like to see an 8+ year promise from them.

Or am I mistaken here?

Only because Apple lost a lawsuit otherwise they'd have kept intentionally slowing it down.

Apple says it made these changes for other reasons, honestly, truly. And if it happened to have the same effect, then that was unfortunate, and unintended.

Only Apple really knows. But there was a slew of changes and reversals following the drama. "Oh, we'll implement notifications now", "Oh, we'll change the peak performance behavior", and "we will change and add additional diagnostics to make sure issues are battery related" certainly has a feel for a bunch of ex post facto rationalization of several things that seem, to me, that if it was truly a battery thing all along, would have been functional requirements.

iPhones are not ubiquitous here, and they're way more expensive than other options.

Who’s at fault for not having built-in messaging that works with rich text, photos, videos, etc?

Google has abandoned more messaging products than I can remember while Apple focused on literally the main function of a phone in the 21st century. And they get shit for it

RCS as currently implemented is iMessage but with a coat of google paint. There is no there there.

Source: I'm an Android user with a lot of iPhones on my in-laws side.

[1] in 2013 they even marketed their “eBeacon” technology as a way for retail stores to monitor and track their customers which…

Privacy wasn’t really a concern because most people didn’t have the privacy eroding device yet. In the years following the Nexus 5 is where smartphones went into geometric growth and the slow realization of the privacy nightmare became apparent

Imho I was really excited to get a Nexus 4 at the time, just a few short years later the shine wore off and I was horrified at the smartphone enabled future. And I have a 40 year background in computers and understand them better than 99 out of 100 users – if I didn’t see it, I can’t blame them either

What a strange statement. I was late to the game with a Nexus S in 2010, and it was really usable.

Until we figure out how to have guarantees of forever, the best we can realistically do is evaluate companies and their products by their behavior now weighted by their behavior in the past.

Somewhat true but things are changing. While there are plenty of “luxury” Apple devices like Vision Pro or fully decked out MacBooks for web browsing we no longer live in a world where tech are just lifestyle gadgets. People spend hours a day on their phones, and often run their life and businesses through it. Even with the $1000+/2-3y price tag, it’s simply not that much given how central role it serves in your life. This is especially true for younger generations who often don't have laptops or desktops at home, and also increasingly in poorer-but-not-poor countries (say eg Eastern Europe). So the iPhone (their best selling product) is far, far, far more a commodity utility than typical luxury consumption like watches, purses, sports cars etc.

Even in the higher end products like the MacBooks you see a lot of professionals (engineers included) who choose it because of its price-performance-value, and who don’t give a shit about luxury. Especially since the M1 launched, where performance and battery life took a giant leap.

It's great having a chip that is so much different than what our production infrastructure uses.

Windows requires a VM, like WSL, for a lot of people, and Linux lacks commercial support. macOS strikes a good balance in the middle that makes it a pretty compelling choice.

Sure, Wine exists, but it’s not something I’d want to rely on for a business when there are alternatives like macOS which will offer native support.

Apple fanboys like to talk about how cool and long lasting a MacBook Air is but a 500 bucks Chromebook will do just as well while allowing pretty much 90% of the use cases. Sure, the top end power is much lower but at the same time considering the base RAM/storage combo Apple gives it is not that relevant. If you starting loading it up, that puts the pricing in an entirely different category and in my opinion the MacBook Air becomes seriously irrelevant when compared to serious computing devices in the same price range...

A Chromebook "does the job" but it's closer to a thin client than a workstation. A lot of the job is done remotely and you may not want that.

If the only thing you care about are battery life (only if you plan to use it lightly on the go, because even the high-end Apple Silicon sucks decent amount of power at full tilt) and privacy I guess they are decent enough.

This is my argument: the base models are at the same time overkill and too limited considering the price and the high-end models are way too expensive for what they bring to the table.

Apple has a big relevancy problem because of how they put a stupid pricing ladder on everything, but that is just my opinion, I guess. As long as they keep making a shit ton of cash it doesn't matter, I suppose. But if the relevant people stop buying Macs because they make no sense, it will become apparent why it matters sooner or later...

And AFAIK, Google dictates 10+h of battery life with mixed web browsing for all Chromebooks.

And if we are talking about expensive high-end hardware, mini-LED is worse than OLED found in those machines anyway so it's not like if that would be a strong argument.

no, i’m a big fan of OLED, but for a laptop, especially a non gaming one, the extra brightness and lack of burn in concern makes it better.

Sure, the MacBook Air is better built and will do some things better but those things are not extremely relevant for someone who would be satisfied by an entry level MacBook Air, because while an MBA has some very nice attributes, in the end everything is limited by its RAM/storage (and to a lesser degree, ports).

For a concrete example, in my country the cheaper MacBook Air you can get is the old M1 design at 1k€, then there is the M2 refresh at 1.2k€ and M3 variant at 1.3k€.

But you can get an Asus Chromebook Plus for 600€ that has either the same amount of RAM and storage or more RAM (16Gb) or more storage (512Gb) depending on the variant you end up with. The display is worse (100 nits less bright and worse resolution) but slightly bigger (14") and that may matter more to many people. It has an older Intel i5 (you can find some AMD options for better efficiency) but it hardly matters for the vast majority of people who just want a laptop to do the basics (basically the target of a MacBook Air). Its battery life would be a bit worse than an MBA but not in a way that can be relevant for the vast majority of customers. One major advantage it has over an MBA is better ports selection, with an HDMI port, a USB A port and an SD card reader on top of the 2 Thunderbolt/USB C ports the MBA has, allowing a dongle free life without having to buy anything else, providing much better utility. That can be way more relevant for many peoples than a better build quality (that I would argue do not even bring better longevity, since with Apple you are hostage of the software support anyway).

You see I am not against MacBooks; in fact, I would advise purchasing a MacBook Pro for some specific use case.

But the reality is that the entry level Apple hardware is rather compromised for its price, and if someone would be satisfied by that choice, I'm arguing that there is another choice (worse on paper, better in some ways) but at half the price (40% off minimum).

If you start loading up a MacBook Air, you end up in MacBook Pro price territory and it doesn't make a lot of sense to not add the 100-200 more to get the much better machine.

I know from experience that entry level Apple hardware is a terrible deal, both because I made the mistake myself or I had to help/fix the issues for other people that made those choices. I have a cousin who remind me every time how much he fucking hates his entry level iMac (an old one with a compromised Fusion Drive and minimum RAM) even though it was rather expensive compared to most computers. My answer is always the same: you spent a lot, but not enough, because with Apple you do not deserve a good experience if you don't go above and beyond in your spending.

In my opinion it is way more disingenuous to advocate for entry-level Apple hardware to people who would be very much satisfied with products costing half as much. The value is just not there, Apple went way too far in the luxury territory in locking down everything and creating a pricing ladder that is engineered to confuse and upsell customers to extract as much money as possible.

For someone who really needs a computer to do more intense work, provided they can work around the tradeoffs of Apple Silicon/macOS and they are willing to spend a large amount of cash, Apple has some great hardware for sure. For everyone else the value is extremely questionnable, especially since they are going full steam ahead into services subscription and the software can be lacking in some ways that will require purchasing even more stuff, the total cost of ownership doesn't make sense anymore. For example, their iPhone SE is absolutely terrible, at 500€ you pay for 6 years old technology with small screen compared to the footprint, terrible battery life, etc. A 500€ mid-range Android is so much better in so many ways that it is just stupid at this point...

As for OLED, I don't think burn-in is a significant concern anymore, and if it is I would argue that you are using it too much like a desktop. In my country you could buy 2 decent OLED laptops for the price of an entry-level MacBook Pro anyway so it doesn't matter as much (and replacing displays of hardware manufacturers other than Apple is much easier and cheaper, so there is that). I think the MacBook Pros are very good for some niche applications, but at viable minimum 2.23k€ price (16Gb RAM/512GB storage) there are a lot of good options so it really requires a careful analysis of actual use case. If you do things related to 3D or engineering it is probably not worth it...

I could see a front end dev needing Photoshop for some things, if they don’t have a design team to give them assets.

There are also security software the company says laptops must have which isn’t available for Linux. They only buy and deploy this stuff with Windows and macOS in mind.

A couple weeks ago on HN I saw someone looking for a program to make a demo of their app (I think). The comments were filled with people recommending an app on macOS that was apparently far and away the best option, and many were disappointed by the lack of availability elsewhere. I find there are a lot of situations like this, where I might be able to get the job done on another OS, but the software I actually want to use is on macOS. Obviously this one is a matter of taste to some degree.

It’s not as big an issue as it was 20 years ago, but it’s still an issue for in many environments.

So this can be done. I don't expect the IT support to help me with any Linux issues.

My excuse for using Linux? It makes me more effective at developing software.

WSL is also nice for Bash scripting, but it's not necessary.

It is a check box in the "Add Features" panel. There is nothing to install or setup. Certainly not for linting, unless, again, you're using a Linux tool chain.

But if you are, just check the box. No setup beyond VS Code, bashrc, vimrc, and your tool chain. Same as you would do on Mac.

If anything, all the Mac specific quirks make setting up the Linux tool chains much harder. At least on WSL the entire directory structure matches Linux out of the box. The tool chains just work.

While some of the documentation is in its infancy, the workflow and versatility of cross platform development on Windows, I think, is unmatched.

I would love to buy Apple hardware, but not from Apple. I mean: M2 13 inch notebook with access to swap/extend memory and storage, regular US keyboard layout and proper desktop Linux (Debian, Alpine, Mint, PopOS!, Fedora Cinamon) or windows. MacOS and the Apple eco system just gets in your way when you're just trying to maintain a multi-platform C++/Java/Rust code base.

Stay classy, Hacker News

Being wrong is no big deal. Being unable to be right is often a very big deal.

Feel free to correct me and share successfull cases with wsl1 and containers

My 15 year old Sony laptop can do better.

Even if Microsoft on Mac is an unmitigated dumpster fire, this is ridiculous.

I avoid using it whenever possible. If people email me, it’d better not be urgent.

Still doesn't take a minute to launch Outlook, but I understand your pain.

I keep hoping it will die, because it would be replaced with an M-series MBP and they are way, way, WAY faster than even the best Intel MBP.

The M1 air was only released in November 2020. With a bit of slop in the numbers, its very possible the parent poster bought an intel mac just before the M1 launched.

It is Microsoft. I could rant all day about the dumpster fire that is the "NEW Microsoft Teams (Work or School)"

It's like the perfect shining example of how MS doesn't give a flaming fuck about their end users.

And while my particular problems might be partially because I am on MacOS, I observe Windows-using colleagues have just as many problems joining meetings (either total refusal, no audio, or sharing issues). So I think using Teams as a measure of any computer is probably not warranted.

(In fact I DO have a spare 13 year old laptop hanging around that still gets used for web browsing, mail and stuff. It is not slow.)

It's fascinating to me how people are still attached to the hardware upgrade cycle as an idea that matters, and yet for a huge chunk of people and scenarios, basically an SSD, 8gb of RAM and an Intel i5 from a decade ago could have been the end of computing history with no real loss to productivity.

I honestly look at people who use Apple or Windows with a bit of pity, because those ecosystems would just give me more stuff to worry about.

It looks like Apple's silicon and software is really good for those doing audio/video. Why people like it for dev is mostly a mystery to me. Though I know a few people who don't really like it but are just intimidated by Linux or just can't handle the small UX differences.

The UX differences are absolutely massive. Even after daily-driving that thinkpad for months, Gnome always felt kinda not quite finished. Maybe KDE is better, but it didn't have Wayland support when I was setting that machine up, which made it a non-starter.

The real killer though is battery life. I can work literally all day unplugged on the mbp and finish up with 40-50% remaining. When i'm traveling these days, i don't even bring a power cable with me during the day. The thinkpad, despite my best efforts with powertop, the most aggressive frequency scaling i could get, and a bunch of other little tricks, lasts 2 hours.

There are niceties about Linux too. Package management is better and the docker experience is _way_ better. Overall though, i'd take the apple silicon macbook 10 times out of 10.

This echoes my experiences for anything that needs power management. Not just that the battery life is worse, but that it degrades quickly. In two years it’s barely usable. I’ve seen this with non-Apple phones and laptops. iPhone otoh is so good these days you don’t need to upgrade until EOL of ~6 years (and even if you need it battery is not more expensive than any other proprietary battery). My last MacBook from 2011 failed a couple of years ago only because of a Radeon GPU inside with a known hw error.

> There are niceties about Linux too.

Yes! If you haven’t tried in years, the Linux desktop experience is awesome (at least close enough) for me – a dev who CAN configure stuff if I need to but find it excruciatingly menial if it isn't related to my core work. It’s really an improvement from a decade ago.

I think the culprit is more likely the power hungry intel CPU in your yoga?

Going on a slight tangent; I've tried but do not like the mac keyboards, they feel very shallow to me, hence why I'm still using my old T480s. The newer thinkpad laptop keyboards all seem to be going that way though (going thinner), much to my dismay. Perhaps a P14s is my next purchase, despite it's bulk.

Anybody with a framework 13 want to comment on their keyboard?

I used a Mac for awhile back in 2015 but it never really stood out to me UX-wise, even compared to Gnome. All I really need to do is open a few windows and then switch between them. In i3 or Sway, opening and switching between windows is very fast and I never have to drag stuff around.

Examples?

My first gen ThinkPad Nano X1 would be an excellent laptop, if it weren’t for the terrible battery life even in power save mode (which as an aside, slows it down a lot) and its need to spin up a fan to do something as trivial as driving a rather pedestrian 2560x1440 60hz display.

It feels almost like priorities are totally upside down for x86 laptop manufacturers. I totally understand and appreciate that there are performance oriented laptops that aren’t supposed to be good with battery life, but there’s no good reason for there being so few ultraportable and midrange x86 laptops that have good battery life and won’t fry your lap or sound like a jet taking off when pushed a little. It’s an endless sea of mediocrity.

My builds are faster, laptop stays cooler, and battery lasts longer. I love it.

If I was building desktop apps I assume it would be a less pleasant experience like you mention.

Modern language tooling (Go, Rust et al) allows one to build and test on any architecture, and the native macOS virtualization (https://developer.apple.com/documentation/virtualization) provides remarkably better performance compared to Docker (which is a better explanation for its fading from daily use).

Your "trend" may, in fact, not actually reflect the reality of how cloud development works at scale.

And I don't know a single macOS developer that "lean(s) on the team member who is on Linux" to leverage tools that are already present on their local machine. My own development environments are IDENTICAL across all three major platforms.

If you deploy on docker (or Kubernetes) on Linux in production, then ideally you should be using docker on your local system as well. Which, for Windows or MacOS users, requires a Linux VM as well.

I know you mean well, though.

There is no Docker on macOS without a hypervisor layer - period - and a VM, though there are multiple possible container runtimes not named Docker that are suitable for devops-y local development deployments (which will always, of course, be constrained in comparison to the scale of lab / staging / production environments). Some of these can better leverage the Rosetta 2 translation layer that Apple provides, than others.

Specifically what confused me in your comment was that you were saying Docker on Mac was superseded by their new native virtualization, which just doesn't make sense to me, for the reasons I was bringing up. I still don't understand what you were trying to say; replacing docker with podman or containerd or something else still doesn't have anything to do with virtualization or Rosetta, or at least I don't see the connection.

I should also say that I don't think anyone really means specifically docker when they talk about it, they probably mean containerization + image repos in general.

There’s quirks to either OS.

Eg when on Gnome it drives me mad that it won’t focus a recently launched apps.

On macOS it annoys me that I have install a 3rd party util to move windows around.

Meh, you just adapt after a while.

Asahi Linux is shockingly great on Apple Silicon hardware, though.

So even if all developers of your team are using Linux, unless you want to waste money by ignoring arm64 instances on cloud computing, you'll have to setup cross compilation.

I actually rejected a job offer when heard I will be given a macbook pro.

Apple, been the most closed company these days, should be avoided as much as you can, not to mention its macos is useless for linux developers like me, anything else is better.

its keyboard is dumb to me(that stupid command/ctrl key difference), can not even mouse-select and paste is enough for me to avoid Macos at all costs.

Probably best for you both.

For what it's worth, I've had a good success rate at politely asking to be given an equivalent laptop I can put linux on, or provide my own device. I've never had to outright reject an offer due to being required to use a Mac. At worst I get "you'll be responsible for making our dev environment work on your setup".

Only one time did I interview with a place where I asked if I'd be given a choice what hardware/OS I could use. The response was "We use Windows". My response was, "no we do not. Either I will not be using Windows with you, or I will not be using Windows NOT with you". I didn't get an offer. I was cool with it.

Now the performance of Mx Macs convinced me to switch, and I'll die on the hill of Android for life

Pretty much everything going wrong with provisioning involves going extra weird on hw (usually for cheap supplier) and/or pushing weird third party "security" crapware.

I don't even know what you mean by mouse-select and paste.

Presumably they mean linux-style text select & paste, which is done by selecting text and then clicking the middle mouse button to paste it (no explicit "copy" command).

macOS doesn't have built-in support for this, but there are some third-party scripts/apps to enable it.

For example: https://github.com/lodestone/macpaste

Then there's Docker. Running amd64 containers on Apple silicon is slow for obvious reasons. Running arm64 containers is fast, but the actual environment you will be deploying to is almost certainly amd64, so if you're using that locally for dev & test purposes, you can get some surprises in prod. Windows, of course, will happily run amd64 natively.

that’s up to your team of course, but graviton is generally cheaper than x86 instances nowadays and afaik the same is true on google and the other clouds.

For most devs, doing crud stuff or writing high-level scripting languages, this isn't really a problem. For some devs, working on time-sensitive problems or with strict baseline performance requirements, this is important. For devs developing device drivers, emulation can only get you so far.

No, I said you won’t always be deploying on amd64. Because arm64 is now the cheapest option and generally faster than the sandy bridge vcpu unit that amd64 instances are indexed against (and really, constrained to, intentionally, by AWS).

I never said anything about graviton not being arm64.

Very quickly remedied by installing the GNU versions of those commands, ie: "brew install coreutils findutils" (etc)

Then you'll have exactly the same command line switches as on Linux.

Literally best keyboard shortcuts out of all major OSes. I don't know what weird crab hands you need to have to comfortably use shortcuts on Windows/Linux. CMD maps PERFECTLY on my thumb.

I.e the same functionality exists with no draw backs and money was no object.

And they chose Windows? Seriously why?

I don't know what to say except resign to the fact that the world is fundamentally unfair, and you won't ever get to run the A/B experiment that you want. So yes, Windows it is !

> where non-Apple monitors and other 3rd party hardware just stops working after minor updates and then starts working again after another update, without any official statement from Apple that there was a problem and a fix?

At least my WiFi doesn't turn off indefinitely during sleep until I power cycle whole laptop because of a shitty driver.

The only problem is that Apple is even worse at communicating issues than Microsoft is.

[1] https://www.bleepingcomputer.com/news/microsoft/microsoft-wa...

[2] https://www.bleepingcomputer.com/news/microsoft/microsoft-fi...

[3] https://www.bleepingcomputer.com/news/microsoft/microsoft-sa...

[4] https://www.forbes.com/sites/gordonkelly/2019/06/12/microsof...

[5] https://www.theregister.com/2022/08/22/windows_10_update_kil...

With Apple, it's usually just crickets... nothing in the release notes, no official statements, nothing. It's just trial and error for the users to see if a particular update fixed the issue.

Not that the degree means much, I learnt 90% of what I know on the job. It certainly helped get my foot in through the university brand, and alumni network.

You can call yourself anything you want Doctor, Lawyer, Engineer. I have the freedom to think my own thoughts too.

There was brief sidestep from late 1980s to early 2010s (~2012) where the term "software engineer" came into vogue and completely ran orthogonal to "proficiency in calculus". I mean, literally 99% of software engineers never learned calculus!

But it's nice to see that ever since ~2015 or so (and perhaps even going forward) proficiency in calculus is rising to the fore. We call those "software engineers" "ML Engineers" nowadays, ehh fine by me. And all those "computers" are not people anymore -- looks like carefully arranged sand (silicon) in metal took over.

I wonder if it's just a matter of time before the carefully-arranged-sand-in-metal form factor will take over the "engineer" role too. One of those Tesla/Figure robots becomes "proficient at calculus" and "proficient at calculations" better than "people".

Reference: [1]: I took the terms "engineer" and "computer" literally out of the movie "Hidden Figures" https://en.wikipedia.org/wiki/Hidden_Figures#Plot

It looks like ever since humankind learned calculus there was an enormous benefit to applying it in the engineering of rockets, aeroplanes, bridges, houses, and eventually "the careful arrangement of sand (silicon)". Literally every one of those jobs required learning calculus at school and applying calculus at work.

Might be just my Eastern Europe background where it was all just "Math" and both equations (that's Algebra I guess) and simpler functions/analysis (Calculus?) are taught in elementary school around age 14 or 15.

Maybe I'm missing/forgetting something - I think I used Calculus more during electrical engineering than for computer/software engineering.

>I think I used Calculus more during electrical engineering than for computer/software engineering.

I think that was OPs point - most engineering disciplines teach it.

It was not really necessary through all of the app developers eras. In fact, it’s so much so the case that many software engineers graduating from 2000-2015 or so work as software engineers without a degree in BS. Rather, they could drop the physics & calculus grind and opt for a BA in computer science. They then went on to become proficient software engineers in the industry.

It’s only after the recent advances of AI around 2012/2015 did a proficiency in calculus become crucial to software engineering again.

I mean, there’s a whole rabbit hole of knowledge on the reason why ML frameworks deal with calculating vector-Jacobian or Jacobian-vector products. Appreciating that and their relation to gradient is necessary to design & debug frameworks like PyTorch or MLX.

Sure, I will concede that a sans-calculus training (BA in Computer Science) can still be sufficiently useful to working as an ML engineer in data analytics, api/services/framework design, infrastructure, systems engineering, and perhaps even inference engineering. But I bet all those people will need to be proficient in calculus the more they have to deal with debugging models.

I don’t mean to disregard other branches of math — of course they’re useful — but calculus stands out in specific _applicability_ to engineering.

Literally every single branch of engineering. All o then. Petrochemical engineering to Biotech. They all use calculus as a fundamental block of study.

Discovering new drugs using Pk/Pd modeling is driven by modeling then drug<->pathogen repo as cycles using Lotka models.

Im not saying engineers dont need to learn stats or arithmetic. IMO those are more fundamental to _all_ fields, janitors or physicians or any field really. But calculus is fundamental to engineering alone.

Perhaps, a begrudging exception I can make is its applications in Finance.

But every other field where people build rockets, cars, airplanes, drugs, or ai robots, you’d need proficiency in calculus just as much as you’d need proficiency in writing or proficiency in arithmetic.

I've come to appreciate the fact that domain knowledge has a more dominant role in solving a problem than technical/programming knowledge. I often wonder how s/w could align with other engineering practices in terms of approach design in a standardized way so we can just churn out code w/o an excessive reliance on quality assurance. I'm really hoping visual programming is going to be the savior here. It might allow SMEs and Domain experts to utilize a visual interface to implement their ideas.

Its interesting how python dominated C/C++ in the case of the NumPy community. One would have assumed C/C++ to be a more a natural fit for performance oriented code. But the domain knowledge overpowered technical knowledge and eventually people started asking funny questions like

https://stackoverflow.com/questions/41365723/why-is-my-pytho...

My strong “opinion” here comes from an observation that while calculus may have been a required subject of study in awarding engineering degrees, the reality is, people didn’t really study it. They just brushed through a couple of pages and wrote a few tests/exams.

In America there’s plethora of expert software engineers who opt for a bachelors degree in computer science that is a BA not a BS.

I think that’s complete totally reasonable thing to do if you don’t want to grind out the physics and calculus courses. They are super hard after all. And let’s face it, all of the _useful to humanity_ work in software doesn’t require expertise in physics or calculus, at least until now.

With AI going forward it’s hard to say. If more of the jobs shift over to model building then yes perhaps a back to basics approach of calculus proficiency could be required.

I guess I just enjoy things more when I can count them.

Calculus was important during the world wars because it means we could throw shells to the enemy army better, and that was an important issue during that period.

Nowadays, calculus is just a stepping stone to more relevant mathematics.

Both are math, and both are still incredibly important. Rockets haven't gone out of style.

Todays ML frameworks grapple with the problem of “jacobian-vector products” & “vector-jacobian product” as a consequence of understanding the interplay between gradients & derivative; and the application of the “chain rule”. All of those 3 concepts are fundamentally understood by being proficient in calculus.

While I’m being the hype-man for calculus I don’t mean to say proficiency in linear algebra or statistics is in any “less necessary” or “less useful” or “less challenging” or “less..” in any way.

I’m merely stating that, historically, calculus has been the unique branch of study for engineering. Statistics has always found value in many fields — business, finance, government policy etc.

Sure Linear algebra is one of those unique fields too — I kinda like to think of it as “algebra” in general and perhaps its utility has flowed in tandem with calculus. Idk. I haven’t thought super hard about it.

When I started doing this "Internet stuff" we were called "webmasters", and job would actually include what today we call: - DevOps - Server/Linux sysadmin - DB admin - Full stack (backend and frontend) engineer

And I might have forgot some things.

There's little things too, like some people using gaming peripherals (multi-button MMO mice and left hand controllers, etc.) for editing, which might not be compatible with OSX.

And also, if you're mucking around with two 32 inch 4k monitors and a 16 inch Wacom it might start to feel a little ridiculous trying to save space with a Mac Pro.

I am not an artist and also dislike multiple monitors, though I will employ two of them on occasion.

My reasons are:

If the window and application management doesn't suck, one display is all one needs.

With cheap multiple displays and touch devices came an ongoing enshitification of app and window management. (And usually dumber focus rules)

Having to turn my head x times a day sucks.

Once it works, Linux is more reliable than Windows. And virtual desktops have always worked better on Linux than on Windows. So I disagree with you on that front.

* In my case, this means I had to get an Anker HDMI adapter, instead of any random brand.

Maybe as long as you don't need working fractional scaling with different DPI monitors, which is nothing fancy now.

I think engineers using software should not be seen as an endorsement. They seem to have an incredible tolerance for bad UI.

People hate software that they're forced to use. Professionals are better at tolerating crapware, because there's usually sunk cost fallacy involved.

Well actually, I use FreeBSD as my daily driver (on a used ThinkPad I bought for 300 euros), and I love it. :D

</painfully earnest nerd>

okay, now you're going to tell me that FreeBSD is in the "software nobody uses" category isn't it?

Disguising it as an "observation" is even more ridiculous.

Nah, I'm good. Believe what you want to believe my friend.

Irelevant.

Firstly, it's a tool, not a social media platform designed to sell ads and farm clicks, it needs to be utilitarian and that's it, like a power drill or a pickup truck, not look pretty since they're not targeting consumers but solving a niche set of engineering problems.

Secondly, the engineers are not the ones paying for that software so their individual tolerance is irelevant since their company pays for the tools and for their tolerance to those tools, being part of the job description and the pay.

Unless you run your own business , you're not gonna turn down lucrative employment because on site they provide BOSCH tools and GM trucks while you personally prefer the UX of Makita and Toyota. If those tools' UX slows down the process and makes the project take longer it's not my problem, my job is to clock in at 9 and clock out at 5, that's it, it's the company's problem to provide the best possible tools for the job, if they can.

> I think engineers using software should not be seen as an endorsement.

Where can I find one of those jobs?

But since you asked, Google is famous for low workloads. Or Microsoft. Or any other old and large slow moving company with lots of money, like IBM, Intel, SAP, ASML, Airbus, DHL, Siemens, manufacturing, aerospace, big pharma, transportation, etc. No bootstrapped "agile" start-ups and scale-ups, or failing companies that need to compete in a race to the bottom.

Depends mostly on where you live though.

They may be locked in, which just forces things. Not an endorsement.

However, they may also be really productive with whatever it is. This could be an endorsement.

In CAD, as an example, there are often very productive interaction models that seem obtuse, or just bad to people learning the tools first time.

Often improving on first time ramp ups to competence nearly always impacts the pro user too.

Where it plays out this way, I have always thought the UI was good in that the pros can work at peak efficiency. It is hard to beat them.

Fact is the task complexity footprint is just large enough to make "good", as in simple, intuitive interfaces not possible.

My MBP has been running without any restart for over a month.

I am not meaning software is engineering or not.

It is a fact, in terms of cost, that software and bridge building are, most of the time very different activities with very different goals and cost-benefit ratios.

All those things count when taking decisions about the level of standardization.

About standards... there are lots also and widely used, from networking to protocols, data transfer formats... with well-known strengths and limitations.

It feels to me that you're bitter or had more than one bad experience. Perhaps you keep working with, or come across, bad Engineers as your generalising is inaccurate.

IETF RFCs soon number over 10K; Java, win32, the Linux kernel syscall API are famous for backward compatibility

not to mention the absurd success of standard libraries of Python, Rust, PHP and certain "standard" projects like Django, React, and ExpressJS

> (how many libraries do you need for arrays and vectors and guis and buttons and text boxes and binary trees and sorting, yada yada?)

considering the design space is enormous and the tradeoffs are not trivial ... it's good to have libraries that fundamentally solve the similar thing but in different context-dependent ways

arguably we are using too many libraries and not enough problem-specific in-situ DSLs (see the result of Alan Kay's research the STEPS project at VPRI - https://news.ycombinator.com/item?id=32966987 )

Many if not most computational problems are fundamentally about data and data transformation under constraints - Throughput, Memory, Latency, etc, etc. And for the situations where the tradeoffs are non-trivial, solving this problem is purely about domain knowledge regarding the nature of the data (video codec data, real-time sensor data, financial data, etc) not about programming expertise.

The various ways to high level architect the overall design in terms of client/server, P2P, distributed vs local, threading model, are, IME are not what I would call crazy complicated. There are standard ways of implementing various variations of the overall design which sadly because of a overall roll-your-own mindset, most devs are reluctant to adopt someone elses design. Part of that is that we don't have a framework of knowledge that allows us to build a library for these designs in our head where we can just pick one thats right for our usecase.

I don't agree with your characterization of the design space as 'enourmous'. I'd say most programmers just need to know a handful of design types because they're not working on high performance, low latency, multi-million endpoint scalable projects where as you say things can get non-trivial.

I'll give a shot at an analogy (I'm hoping the nitpickers are out to lunch). The design space for door knob is enormous because of the various hand shapes, disability constraints, door sizes, applications, security implications, etc. And yet we've standardize d on a few door knob types for most homes which you can go out and buy and install yourself. The special case bank vaults and prisons and other domains solve it their own way.

I worked for one of the UKs leading architecture / construction firms writing software and also am an amature mechanic.

You'd be amazed at how many gasket types, nuts, bolts, fasteners, unfasters, glues, concretes, bonding agents and so on ... all invented for edge preferences and most of which could be used interchangably.

Also standards? Hah. They're an absolute shitshow in any engineering effort.

I mean ... even just units of measure. C'mon.

I've set up my OS exactly as I want it. (I use arch btw ;-))

Obviously doesn’t apply to all engineers.

I do networking stuff and macOS is on par with Windows - I can't live on it without running into bugs or very questionable behavior for longer than a week. Same as Windows.

> I do networking stuff

Me too, but probably very different stuff. I’m doing p2p stuff over tcp and am affected mostly by sock options, buffer sizes, tcp options etc.

I like apple hardware, but their OS is fucking atrocious. In the year 2024 it still doesn't have a native volume mixer, or any kind of sensible window management shortcuts. Half the things on it have to be fixed with paid software. Complete joke of an OS, if it were up to me I'd stick a linux distro on top of the hardware and be happy

I think they don't because they have an incentive to not do so: they get a cut on all the software you have to purchase on the App Store to make up for it. It might not look like a lot, but if a large portions of Mac users need to buy a 5-10 bucks app to fix the windows management problems, it becomes serious money at 15-30% cut on millions of purchases...

And this is precisely the problem with Apple today. They are not honest enough to fix or improve the stuff they sell at a very high price, both because they sell it anyway and because they put in place many incentives for themselves to not do so.

There is the running meme of the iPad calculator but macOS could also use some care on the calculator/grapher not having received serious attention in decades. At the price they sell their stuff that would seem like a given, but considering they'll make money on the apps you buy to improve that situation, they'll never do so.

After using Apple App Stores for so many years, I wish I didn't, the convenience really isn't worth the cost down the road...

I know engineers from a FANG that picked MacBook pros in spite of the specs and only because of the bling/price tag. Them they spent their whole time using it as a remote terminal for Linux servers, and they still complained about the thing being extremely short on RAM and HD.

One of them even tried to convince their managers to give the vision pro a try, even though there was zero use cases for it.

Granted, they drive multiple monitors well with a single USB-C plug, at least with specific combinations of monitors and hubs.

It's high time that the "Apple sells high end gear" shtick is put to rest. Even their macOS treadmill is becoming tiring.

Yes, there are 2k+ laptops from Dell/Lenovo that match and exceed a similarly priced MacBook in pure power, but usually lack battery life and/or build quality.

I can't do that on Windows without either wrestling with PowerShell or WSL2

in case Apple would come up with the idea to take an ipad as external compute unit that would be amazing... just double your ram, compute and screen with it in such a lightweight form factor... should be possible if they want

Also the nice thing for headphones within apple is, that the airpods automatically switch to where the attention is... meaning e.g., in case I watch something on the laptop and pick up an iphone call (no matter if via phone or any app) the airpod automatically switches

I owned a MacBook Pro with the dreaded butterfly keyboard. It was shit.

How many USB ports do the new MacBook air have? The old ones had two. And shipped with 8GB of RAM? These are shit-tier specs.

The 2020 MacBook pros had a nice thing: USB-C charging, and you could charge the from either side. Current models went back to MagSafe, only on one side. The number of USB ports is still very low.

But the are shiny. I guess that counts as quality.

I agree, the butterfly keyboard was shitty but I absolutely love the keyboard on the 2020 rev. It's still not as great as my mechanical desktop keyboard, but for a laptop keyboard it's serious chef's kiss. Also, I have yet to find a trackpad that is anywhere as good as the Macbook. Precision trackpads are still way way worse.

Finally, the thing that always beings me back to MBPs (vs Surfacebooks or Razers) is battery life. I typically get a good 10+ hours on my MBP. Battery life on my old Razer Blade and Surfacebooks were absolutely comically horrible.

However there are two awesome things about my work MBP I would really want from my ThinkPad:

Magsafe charger - too many close calls!

And the track pad.

I can't work properly without an external mouse on my ThinkPad. But on the MBP everything just has the right size, location, proportions and handling on the track pad. I had a mouse for the MBP too but I stopped using it!

Yea, the butterfly was utter shit. And they fucked up the touchbar by not just putting it on TOP of the existing F-keys.

But the rest of the laptop was still well machined :D

I am the one paying for my MacBook Pro, because my company is a self-funded business. I run my entire business on this machine and I love it. I always buy the fastest CPU possible, although I don't max out the RAM and SSD.

Amusingly enough, I talked to someone recently about compilation speeds and that person asked my why I don't compile my software (Clojure and ClojureScript) on "powerful cloud servers". Well, according to Geekbench, which always correlates very well with my compilation speeds, there are very few CPUs out there that can beat my M3 Max, and those aren't easily rentable as bare-metal cloud servers. Any virtual server will be slower.

So please, don't repeat the "MacBooks are for spoiled people who don't have to pay for them" trope. There are people for whom this is simply the best machine for the job at hand.

Incidentally, I checked my financials: a 16" MBP with M3 and 64GB RAM, amortized over 18 months (very short!) comes out to around $150/month. That is not expensive at all for your main development machine that you run your business on!

Which, incidentally, is probably about 10x less than you would spend compiling your software on "powerful cloud servers". :-)

I mean Hetzner has a reputation for renting bare metal servers at the cheapest price in the market. Try AX102 which has very close performance to a M3 Max (CPU only): https://www.hetzner.com/dedicated-rootserver/matrix-ax/

While the OP's solution has a lot of advantages like being able to own the device and including GPU, but at least we do have cloud servers with comparable costs available.

Worth noting: the monthly cost is close to my 18-month amortized cost.

I use Debian + FreeBSD on my personal PCs though.

For dev (at least backend and devops), macOS is not that great.

Still, I just don't think it's that burdensome to get containers running on MacOS, it's just annoying that it happens to work worse than on Windows or Linux. Ignoring the hardware, the only real advantage to MacOS development is when you're targeting Apple products with what you're developing.

Not all teams run tons of containers on personal computers. All our servers are running on AWS. I rarely ssh into anything.

I like the fact that OS X is based on UNIX, and not some half-assed bullshit bolted onto Windows. I still have bad memories of trying to use Cygwin 15 years ago. Apparently WSL is an improvement, but I don't care.

Mac runs all the software I need, and it has real UNIX shells.

The vast majority of the software engineers in big companies (that employ a lot more people than big tech and startups combined) who use Java and C# also have predominately Windows laptops (as their employers can manage Windows laptops a lot easier, have agreements with vendors like Dell to buy them with a discount, have software like AV that doesn't support MacOS, etc.).

On top of that MacBooks don't have the best screens and are not the best built. Many Windows laptops have OLED screens or 4K IPS screens. There are premium Windows laptops made out of magnesium and carbon fiber.

Every company I've worked for during the last 12 years gives out MacBook Pros. And I've been developing using Scala / Java for the last 20 years.

Employers manage Macs just fine, this isn't 1999. There have been studies showing that Macs have lower IT maintenance costs compared to Windows.

I admit that I haven't dealt with Windows devices in a long time, maybe there are some good ones available now, but I find your statements to be beyond belief. Apple Silicon Macs have blown the doors off the competition, out performing all but top-end Intel laptops, while using a fraction of the power (and I never even hear the fans come on).

I am happy to not support Apple's ecosystem and use a minimally worse laptop from a different brand.

The same is true of the ipad which is just a miraculous piece of hardware constrained by an impotent operating system.

Apple is selling hardware and scaling AI by utilizing it is simply a smart move.

Instead of building huge GPU clusters, having to deal with NVIDIA for GOUs (Apple kicked NVIDIA out years ago because of disagreements), Apple is building mainly on existing hardware.

This is in other terms utilizing CPU power.

On the other hand this helps their marketing keeping high price points when Apple now is going to differentiate their COU power and therefore hardware prices over AI functionality correlating with CPU power. This is also consistent with Apple stopping the MHz comparisons years ago.

Also, Siri, and consider: you’re scaling AI on apple’s hardware, too, you can develop your own local custom AI on it, there’s more memory available for linear algebra in a maxed out MBP than the biggest GPUs you can buy.

They scale the VRAM capacity with unified memory and that plus a ton of software is enough to make the Apple stuff plenty competitive with the corresponding NVIDIA stuff for the specific task of running big AI models locally.

But this hardly applies to 95% if not more people of all people running Apple's hardware, the fastest CPU/GPU isn't worth much if you can fit any at least marginally useful LLM model on the 8GB (or less on iPhones/iPads) of memory that you device has?

Heck now I see even Vietnam iPhone is #1 vendor with a 28% market penetration according to statcounter. That’s more than I thought, even though I was just there…

Speaking of India, they’re at 4% there. That’s closer to being luxury.

> doesn't mind using WhatsApp for messaging

Well WhatsApp was super early and way ahead of any competition, and the countries where it penetrated had no reason to leave, so it’s not exactly like they settle for less. It has been a consistently great service (in the category of proprietary messaging apps), even after Zuck took over.

The cheap Windows PCs suck just like a cheap car sucks (ok, they suck more), but they still get the job done. You can still browse the web, read your email, watch a youtube video, post a youtube video, write a blog, etc.. My dad got some HP celeron. It took 4 minutes to boot. It still ran though and he paid probably $300 for it vs $999 for a mac. He didn't have $999.

> The cheap Windows PCs suck […], but they still get the job done

For desktop, totally. Although I would still wipe it with Ubuntu or so because Windows is so horrible these days even my mom is having a shit time with only browsing and video calls.

A random laptop however is a different story. Except for premium brands (closer to Apple prices) they tend to have garbage battery life, infuriating track pad, massive thermal issues, and preloaded with bloatware. Apple was always better here, but now with the lower power/heat of the ARM chips, they got soooo much better overnight.

To the person with no budget, all that doesn't matter. They'll still get let $250 laptop and put up with the garbage battery life (find a power outlet), infuriating trackpad (buy an external mouse for $10), bloatware (most users don't know this and just put up with it), etc....

I agree Apple is better. But if your budget is $250 and not $1k then you get what you can get for $250 and continue to feed your kids and pay your rent.

I notice your use of the weasel word "just".

We undoubtedly live in a world where Apple products are sold as lifestyle gadgets. Arguably it's more true today than it ever was. It's also a world where Apple's range of Veblen goods managed to gain footing in social circles to an extent that we have kids being bullied for owning Android phones.

Apple's lifestyle angle is becoming specially relevant because they can no longer claim they sell high-end hardware, as the difference in specs between Apple's hardware and product ranges from other OEMs is no longer noticeable. Apple's laughable insistence on shipping laptops with 8GB of RAM is a good example.

> Even in the higher end products like the MacBooks you see a lot of professionals (engineers included) who choose it because of its price-performance-value, and who don’t give a shit about luxury.

I don't think so, and that contrasts with my personal experience. All my previous roles offered a mix of MacBooks and windows laptops, and MacBooks were opted by new arrivals because they were seen as perks and the particular choice of windows ones in comparison were not as impressive, even though they out-specced Apple's offering (mid-range HP and Dell). In fact in a recent employee's review their main feedback was that the MacBook pro line was under-specced because at best it shipped with only 16GB of RAM while the less impressive HP ones already came with 32GB. In previous years, they called for the replacement of the MacBook line due to the rate of keyboard malfunctions. Meaning, engineers were purposely picking the underperforming option for non-technical reasons.

Also, I think you’re misunderstanding what a Veblen good is and the difference between “premium” and “luxury.” Apple does not create luxury or “Veblen” goods like for example, LVMH.

An easy way to discern the difference between premium and luxury — does the company advertise the product’s features or price?

For example, a Chanel handbag is almost entirely divorced from its utility as a handbag. Chanel doesn’t advertise features or pricing, because it’s not about the product’s value or utility, it’s what it says about your personal wealth that you bought it. That’s a Veblen good.

Apple heavily advertises features and pricing. Because they sell premium products that are not divorced from their utility or value.

I own a MacBook Air, I won't be buying another purely because the moment I need to upgrade anything or repair anything it's effectively ewaste.

Fiddler on windows works flawlessly.

Most CS professionals who write code have no idea what it takes to build a desktop, so the hardware that they chose is pretty much irrelevant because they aren't specifically choosing for hardware. The reason Apple gets bought is mostly by anyone, including tech people, is because of ecosystem. The truth is, nobody really care that much about actual specs as long as its good enough to do basic stuff, and when you are indifferent to the actual difference but all your friends are in the ecosystem, the choice is obvious.

You can easily see this yourself: ask these "professionals" about the details of the Apple Neural engine, and its a very high chance that they will repeat some marketing material, while failing to mention that Apple does not publish any real docs for ANE, you have to sign your code to run on ANE, and you have to basically use Core ML to utilize the ANE. I.e if they really cared about inference, all of them would be buying laptops with discrete 4090s for almost the same price.

Meanwhile, if you look at people who came from EE/ECE (who btw on the average are far better coders than people with CS background, based on my 500+ interviews in the industry across several sectors), you see a way larger skew towards Android/custom built desktops/windows laptops running Linux. If you lived and breathed Linux and low level OS, you tend appreciate all the power and customization that it gives you because you don't have to go learn how to do things.

The signing of binaries is a part of adult developer life, and is certainly required for the platforms you mention as well.

Unquestionably, battery life on 4090-based laptops sucks on a good day, and if you're working long hours, the last thing you want to have to do is park yourself next to your 350W adapter just to get basic work done.

Very much not true. Not to make this personal, but this is exactly what Im talking about Apple fans not understanding hardware.

Linux has been through the ringer of fighting its way to general use, and because of its open source nature and constant development. So in terms of working well, it has been optimized for hardware WAY further than Apple, which is why you find it on servers, personal desktops, phones, portable gaming devices, and even STM32 Cortex bldc control boards, all of which run different hardware.

Apple doesn't optimize for general use, it optimizes for a specific business case. In the case of Apple silicon, it was purely battery life which brings more people in to the ecosystem. Single core performance is on par with all the other chips, because the instruction set doesn't actually matter (https://chipsandcheese.com/2021/07/13/arm-or-x86-isa-doesnt-...), multi core is behind, Mac Os software is still a pile of junk (Rosetta still isn't good across the board), computers are not repairable, you have no privacy since Apple collects a shitload of telemetry for themselves, e.t.c and so on.

And, Apple has no incentive to make any of this better - prior to Apple Silicon, people were still buying Intel Macs with worse specs and performance for the same price, all for the ecosystem and vanity. And not only was the Macos still terrible (and much slower), you also had hardware failures like plugging in a wrong USBC hub would blow the chip and brick your Mac, butterfly keyboards failing, and questionable decisions like virtual esc keys.

>The signing of binaries is a part of adult developer life,

...for professional use, and the private key holder should be the person who wrote that software. I hope you understand how ridiculous it is to ask a developer to sign code using the manufacturers key to allow them to run that code on a machine that they own.

>Unquestionably, battery life on 4090-based laptops sucks on a good day,

Well yea, but you are not buying that laptop for battery life. Also, with Ryzen cpus and 4090s, most get like 6-8 hours depending on use due to Nvidia Prime, which is pretty good for travel, especially if you have a backpack with a charging brick.

If you want portability, there are plenty of lighter weight option like Lenovo Yoga which can get 11-12 hours of battery life for things like web browsing.

Apple is popular in the US because a) luxury brands hold sway b) they goad customers into bullying non-customers (blue/green chats) and c) they limit features and customizability in favor of simpler interfaces.

It's popular with developers because a) performance is valuable even at Apples steep cost b) it's Unix-based unlike Windows so shares more with the Linux systems most engineers are targeting.

Here I feel that I-hate-apple crowd is just stuck with this notion of luxury overpriced brand when it is clearly not the case. Apple has superior hardware at better price points. Last time I was doing shopping for a laptop, I could get similar features only at a 30% - 40% price premium in other brands.

I am typing this on an apple M2 air and try finding similar performance under 2000 USD in other brands. The responsiveness, the (mostly) sane defaults and superior rendering and fonts make it worth it. The OS does not matter so much as it used to do in 2004 and the fact that I have a unix terminal in 2024 is just incidental. I have turned off auto updates and I do not use much of phone integration apart from taking backups and photo copying.

I switched to an iPhone in 2022 from a 200 US$ Samsung handset. Here, I would say that not everyone needs an iPhone. My old phone used to do all the tricks I need on this one. However, the camera is really and photos are really great. If I buy an iPhone next time, it would be just for the photos it takes.

(FWIW my opinion is it is a human right, I just think Tim Cook is full of shit.)

What Apple calls privacy more often than not is just putting lipstick on the pig that is their anticompetitive walled garden.

Pretty much everybody in SV who works in privacy rolls their eyes at Apple. They talk a big game but they are as full of shit as Meta and Google - and there’s receipts to prove it thanks to this DoJ case.

Apple want to sell high end hardware. On-device computation is a better user experience, hands down.

That said, Siri is utter dogshit so on-device dogshit is just faster dogshit.

The idea that user behavior or consumer choice will change any of this is basically discredited in practice. It will always been cat and mouse until the point that CEOs go to jail, then it will stop.

How is something that is twice as expensive as the competition not a luxury device?

EDIT: Because there's repeated confusion in the replies: I am not saying that a MacBook Air is not objectively a better device. I'm saying it is better by metrics that fall strictly into the "luxury" category.

Better build quality, system-on-a-chip, better OS, better battery life, aluminum case—all of these are luxury characteristics that someone who is looking for a functional device that meets their needs at a decent price won't have as dealbreakers.

Most people, when given the pitch you just gave me for a 2x increase in price, will choose the cheaper item, just like they choose the cheaper car.

[edit] I mean sure there are people who just want the Apple logo, I’m not saying there are zero of those, but they’re also excellent, reliable tools (by the standards of computers—so, still bad) and a good chunk of their buyers are there for that. Even the ones who only have a phone.

The equivalent from Apple would currently run me $3200. If I'm willing to compromise to 24GB of RAM I can get one for $2200.

What makes an Apple device a luxury item isn't that it's more expensive, it's that no matter what specs you pick it will always be much more expensive than equivalent specs from a non-luxury provider. The things that Apple provides are not the headline stats that matter for a tool-user, they're luxury properties that don't actually matter to most people.

Note that there's nothing wrong with buying a luxury item! It's entirely unsurprising that most people on HN looking at the latest M4 chip prefer luxury computers, and that's fine!

The devices are great and feel great. Definitely high quality (arguably, luxury!). The OS leaves a lot to be desired for me.

Got a job in dual-platform mobile dev and was issued a MacBook. Exposure to dozens of phones and tablets from both ecosystem. I was converted within a year.

(I barely customize anything these days, fwiw—hit the toggle for “caps as an extra ctrl”, brew install spectacle, done. Used to have opinions about my graphical login manager, use custom icon sets, all that stuff)

On the phone side, I guess you would call Samsung and Google luxury providers? On the laptop side there are a number of differentiating features that are of general interest.

> The things that Apple provides are not the headline stats that matter for a tool-user, they're luxury properties that don't actually matter to most people

Things that might matter to regular people (and tool users):

- design and build for something you use all day

- mic and speakers that don't sound like garbage (very noticeable and relevant in the zoom/hybrid work era)

- excellent display

- excellent battery life

- seamless integration with iPhone, iPad, AirPods

- whole widget: fewer headaches vs. Windows (ymmv); better app consistency vs. Linux

- in-person service/support at Apple stores

It's hard to argue that Apple didn't reset expectations for laptop battery life (and fanless performance) with the M1 MacBook Air. If Ryzen has caught up, then competition is a good thing for all of us (maybe not intel though...) In general Apple isn't bleeding edge, but they innovate with high quality, very usable implementations (wi-fi (1999), gigabit ethernet (2001), modern MacBook Pro design (2001), "air"/ultrabook form factors (2008), thunderbolt (2011), "retina" display and standard ssd (2012), usb-c (2016), M1: SoC/SiP/unified memory/ARM/asymmetric cores/neural engine/power efficiency/battery life (2020) ...and occasionally with dubious features like the touchbar and butterfly keyboard (2016).)

How much does it weight? Battery life? Screen quality? Keyboard? Speakers?

Here lies the rub, ARE those the stats that matter? Or does the screen, touchpad, speakers, battery life, software, support services, etc. matter more?

I feel people just TOTALLY gloss over the fact that Apple is crushing the competition in terms of trackpads + speakers + battery life, which are hardly irrelevant parts of most people's computing experience. Many people hardly use their computers to compute - they mostly use them to input and display information. For such users, memory capacity and processing performance ARE frills, and Apple is a market leader where it's delivering value.

Also even in compute, apple is selling computers with a 512-bit or 1024-bit LPDDR5x bus for a lower price than you can get from the competition. Apple is also frequently leading the pack in terms of compute/watt. This has more niche appeal, but I've seen people buy Apple to run LLM inferencing 24/7 while the Mac Studio sips power.

Im a Linux guy but am forced to use Mac's and windows every now and then.

Windows has outpaced macos for a decade straight.

Macos looks like it hasn't been updated in years. It's constantly bugging me for passwords for random things. It is objectively the worst OS. I'd rather work on a Chromebook.

Maybe the only reason why Windows outpaced macos in this for you is because Windows started as crap and is now barely looks like a proper OS?

https://ntdev.blog/2021/02/06/state-of-the-windows-how-many-...

Thank God Mac hasn't changed that much. I absolutely love its UI.

A good deal on a luxury item still gets you a luxury item.

And if we want to compare Walmart to Walmart, this thing currently runs for $359 and has 16GB RAM, 512GB SSD, and a CPU that benchmarks slightly faster than the M2:

https://www.walmart.com/ip/Acer-Aspire-3-15-6-inch-Laptop-AM...

I think you unjustifiably downplay how much of a selling point a screen that looks great (or at least decent) on the floor is. And I know tons of devs that put up with the 45% NTSC abominations on Thinkpads that aren’t even suitable for casual photo editing or web media, just because you make do with that doesn’t automatically make a halfway decent display on a laptop a “luxury”.

Sorry, but don’t buy the “everything that isn’t a $300 econo shit laptop is luxury” thesis repeated ad nauseum.

But the definition of luxury varies a lot by product category. Still, high-end and luxury are separate concepts, which ven when they do overlap.

It is true though that luxury goods are, often, top of the line as well. Cars and watches are often examples of this. Clothes are a much more mixed bag, with some luxury brands using excellent materials and craftsmanship, while others use flashy design and branding with mediocre materials and craftsmanship.

Exactly where Apple sits is very debatable in my experience. I would personally say that many of their products are far too affordable and simple to be considered luxury products - the iPhone in particular. The laptops I'm less sure about.

This comes up so often in these flame wars where people are really saying “I do/don’t think you need that feature” and won’t accept that other people aren’t starting from the same point. I remember in the 90s reading some dude on Fidonet arguing that Macs were overpriced because they had unnecessary frills like sound cards and color displays; I wasn’t a Mac user then but still knew this was not a persuasive argument.

I'm seeing significant differences in the performance between Acer Aspire A315 to M2 Macbook Air; the Acer is ~33% of the M2 for 50% the price.

https://browser.geekbench.com/v6/cpu/compare/5942766?baselin...

$999 amortized over 3 years is $30/mo which is less than what even middle class people spend on coffee.

> 3.92 lb

Battery life

> 6.5 h

Probably half of that.

Fans that sound like jet engine, screen quality which would force me to stab my eyes, speakers sounding worse than a toilet, plastic build.

I'm not convinced.

> https://news.ycombinator.com/item?id=40292804

tl;dr is that Walmart is also selling an Acer for $359 that beats that device on every headline metric.

It's nice to know that I could get the old-gen model for slightly cheaper, but that's still an outrageous price if the MacBook Air isn't to be considered a luxury item.

8gb of ram was pathetic in 2018, and is SUPER pathetic in 2024.

These products may not offer you much value and you don't have to buy them. Clearly plenty of people and institutions bought them because they believed they offered the best value to them.

https://www.youtube.com/watch?v=kHKIcBWbnjo

Try this:

https://news.ycombinator.com/item?id=40297295

No one calls a Toyota Camry with base options luxury but it works well for a long time and has good quality.

The features that Apple adds on top of that are strictly optional. You can very much prefer them and think that they're essential, but that doesn't make it so. Some people feel that way about leather seats.

Closer to $999 then $500.

Nobody is away from a power source for longer than 18 hours. MOST people don't need the performance that a macbook air has, their NEEDS would be met by a raspberry pi... that is, basic finances, logging into various services, online banking, things that first world citizens "rely" on.

The definition of luxury is "great comfort and extravagance", and every current Apple product fits that definition. Past Apple definitely had non-luxury products, as recently as the iPhone C (discontinued 10 years ago)... but Apple has eliminated all low-value options from their lineup.

https://www.amazon.com/Lenovo-IdeaPad-Ryzen5-5500U-1920x1080...

An 18 hour battery life is a luxury characteristic, not something penny pinchers will typically be selecting on.

Apple likes to overcharge for storage, but the drives are _really_ good.

As I said in another comment:

The point isn't that the MacBook Air isn't better by some metrics than PC laptops. A Rolls-Royce is "better" by certain metrics than a Toyota, too. What makes a device luxury is if it costs substantially more than competing products that the average person would consider a valid replacement.

Except Apple wanted $3,000 for 7TB of SSD (considering the base price already included 1TB).

Instead, I bought a 4xM.2 PCI card, and 4 2TB Samsung Pro SSDs.

I paid $1,300 for it, got to keep the 1TB "system" SSD.

And I get faster speeds from it, 6.8GBps versus 5.5GBps off the system drive.

For $2,000 I could have got the PCI 4.0 version and SSDs, and get 26GBps.

... for an eye watering $3,000 over the exact same spec Mac Studio.

I liked my cheesegrater, though I didn't like the heat output.

And I cannot justify that. I sacrificed half the throughput (2800MBps) for $379 and got an external 4 x M.2 TB3 enclosure.

Oh, and a USB 3 hub to replace one I had installed in the cheesegrater to augment the built in ports. $400 give or take.

Like I say Audi is a luxury car because a Toyota costs less than half as much, and you ask "what about a Toyota with leather seats"?

My experiences (multiple) with Genius Bar have been decidedly more "meh" to outright frustrating, versus "luxury", oftentimes where I know more than the Genius.

Logic Board issues where on a brand new macOS install I could reproducibly cause a kernel panic around graphics hardware. There was an open recall (finally, after waiting MONTHS) on this. It covered my Mac. But because it passed their diagnostic tool, they would only offer to replace the board on a time and materials basis.

I had a screen delamination issue. "It's not that bad - you can't see it when the screen is on, and you have to look for it". Huh. Great "luxury" experience.

And then the multiple "we are going to price this so outrageously, and use that as an excuse to try to upsell". Like the MBA that wouldn't charge due to a circuit issue. Battery fine, healthy. Laptop, fine, healthy, on AC. Just couldn't deliver current to the battery. Me, thinking sure, $300ish maybe with a little effort.

"That's going to be $899 to repair. That's only $100 less than a new MBA, maybe we should take a look at some of the new models?" Uh, no. I'm not paying $900 for a laptop that spends 99% (well, 100% now) of its life on AC power.

M-series Macs is when MacBooks stopped sucking at battery life without sleeping and wrecking state every moment they could.

https://www.amazon.com/HP-Pavilion-i7-11370H-Micro-Edge-Anti...

The point isn't that the MacBook Air isn't better by some metrics than PC laptops. A Rolls-Royce is "better" by certain metrics than a Toyota, too. What makes a device luxury is if it costs substantially more than competing products that the average person would consider a valid replacement.

MacBook Air (2022): https://browser.geekbench.com/macs/macbook-air-2022

Ryzen 5 5500U (CPU): https://browser.geekbench.com/processors/amd-ryzen-5-5500u

Ryzen 5 5500U (APU, similar laptop): https://browser.geekbench.com/v5/compute/6751456

But what is that, exactly? If you look at all aspects of a laptop: CPU, RAM, SSD, battery life, screen quality, build quality, touchpad, OS, and put them in order of importance for the average consumer, what would be on top? I don't think it's the tech specs.

For instance, I would be willing to bet that for a large number of consumers, battery life is far more important than the tech specs, which means that a valid replacement for their MacBook must have equivalent battery life. You also have to consider things like the expected lifespan of the laptop and its resale value to properly compare their costs. It's not simple.

You can buy a version of <insert product here> from Walmart at 1/2 price of a "normal" retailer. Does that mean every "normal" retailer is actually a luxury goods dealer?

Is my diner a luxury restaurant because a burger costs twice as much as McDonald's?

Stop the silliness.

So for me, yes, MacBook Airs, like the lovely M1 I'm writing this on (I love my job!) are luxury goods.

Just because Ferraris cost half a million doesn't mean a 50k BMW isn't luxury.

That's for the 8GB model (which is already too low unless you use your Mac for the same use cases as you would use an iPad), 16GB is $200 extra.

What percent of that retailer's products does that comparison apply to?

If it's more than half then yeah that's probably a luxury goods dealer.

Equivalent device?! Find me Windows laptop in ANY price category that can match weight, fanless design, screen quality, battery life, speakers quality and battery life of Air.

Cost comparisons alone are stupid. And “this AMD benchmarks the same as an M2” is a useless comparison since regular people don’t buy laptops for raw compute power.

I have used both. I think the Mac experience is significantly better. No one is looking at me.

> Their primary business goal is to sell hardware.

There is no contradiction here. No need for luxury. Efficient hardware scales, Moore's law has just been rewritten, not defeated.

Power efficiency combined with shared and extremely fast RAM, it is still a formula for success as long as they are able to deliver.

By the way, M-series MacBooks have crossed bargain territory by now compared to WinTel in some specific (but large) niches, e.g. the M2 Air.

They are still technically superior in power efficiency and still competitive in performance in many common uses, be it traditional media decoding and processing, GPU-heavy tasks (including AI), single-core performance...

By the way, this includes web technologies / JS.

If you're into AI, there's objectively literally no other laptop on the planet that is competitive with the GPU memory available on an MBP.

If you don't explicitly want a laptop, a 32GB M2 Pro Mac Mini would be a good choice I think.

Personally i only have used MBPs so far.

But the M-series Air are not remotely comparable to the old Intel Airs, that's for sure :)

Tim Cook unfortunately is so captured in that quarterly mindset of 'please the share holders' that it is only a matter of time.

What matters to me is that they continue to see privacy as something they can sell in order to make money.

I do hope that those working in these companies actually building the tools do care. But unfortunately, it seems that corruption is an emergent property of complexity.

Apple is in the position to build a competing search product, but the amount Google pays is the amount of money they would have to earn from it, and that is improbable even if it means they can set their own search engine as default.

The important is that a good starting ai learning platform is what … most apple does not touch those price.

Hence with privacy it is a good path.

You do not want communism even if it is not expensive in the short term.

They failed with their ad-business so this is a nice pivot. I'll take it, I'm not usually a cheerleader for Apple, but I'll support anyone who can erode Google's surveillance dominance.

I have very little faith in apple in this respect.

For clarity, just install little snitch on your machine, and watch what happens with your system. Even without being signed in with an apple id and everything turned off, apple phones home all the time.

iCloud, App store revenue, apple tv, and so on

Everyone seems to have forgotten about the Celebrity iCloud photo leak.

While Apple is first and foremost a hardware company, it has more or less always been about the "Apple experience". They've never "just" been a hardware company.

For as long as Apple has existed, they've done things "their way" both with hardware and software, though they tend to want to abstract the software away.

If it was merely a question of selling hardware, why does iCloud exist ? or AppleTV+, or Handoff ? or iMessage, or the countless other seemingly small life improvements that somehow the remainder of the industry cannot seem to figure out how to do well.

Just a "simple" thing as switching headphones seamlessly between devices is something i no longer think about, it just happens, and it takes a trip with a Windows computer and a regular bluetooth headset to remind me how things used to be.

As part of their "privacy first" strategy, iMessage also fits in nicely. Apple doesn't have to operate a huge instant messaging network, which undoubtedly is not making a profit, but they do, because having one entry to secure, encrypted communication fits well with the Apple Experience. iMessage did so well at abstracting the ugly details of encryption that few people even think about that that's what the blue bubble is actually about, it more or less only means your message is end to end encrypted. As a side effect you can also send full resolution images (and more), but that's in no way unique to iMessage.

Everybody so quick to forget Apple was/is part of PRISM like any other company.

Indeed.

Privacy starts with architectural fundamentals that are very difficult to retrofit...

If a supplier of products has not built the products this way, it would be naive to bet bank or farm on the supplier. Even if there were profound motivation to retrofit.

Add to this the general tendency of the market to exploit its customers.

There are a ton of us out here that consciously choose Apple because of their position on privacy. I have to imagine they know how many customers they'll lose if they ever move on this, and I want to believe that it's a large enough percentage to prevent it from happening. Certainly my circle is not a useful sample, but the Apple people in it are almost all Apple people because of privacy.

E.g. Apple still doesn't offer a flipphone that turns into a tablet when you open it.

Yep, the claim to prioritize privacy is a key selling point for them

(assuming companies won't easily share all their models for this kind of effort)

I'm curious: is anyone seriously using apple hardware to train Ai models at the moment? Obviously not the big players, but I imagine it might be a viable option for Ai engineers in smaller, less ambitious companies.

"No, we should be use cheap commodity abundantly available cpus and orchestrate then behind cloud magic to write our nl translation apps"

or maybe "no we should build purpose built high performance computing hardware to write our nl translation apps"

Or perhaps in the early 70s "is anyone seriously considering personal computer hardware to ...". "no, we should just buy IBM mainframes ..."

I don't know. Im probably super biased. I like the idea of all this training work breaking the shackles of cloud/mainframe/servers/off-end-user-device and migrating to run on peoples devices. It feels "democratic".

For example, running a local model and access to the features of a larger more capable/cloud model are two completely different features therefore there is no "no we should do x instead".

I'd imagine that a dumber local model runs and defers to cloud model when it needs to/if user has allowed it to go to cloud. Apple could not compete on "our models run locally privacy is a bankable feature" alone imo, TikTok install base has shown us enough that users prefer content/features over privacy, they'll definitely still need SoA cloud based models to compete.

An older example is the “Hey Siri” model, which is fine tuned to your specific voice.

But with regards to on device training, I don’t think anyone is seriously looking at training a model from scratch on device, that doesn’t make much sense. But taking models and fine tuning them to specific users makes a whole ton of sense, and an obvious approach to producing “personal” AI assistants.

[1] https://support.apple.com/en-us/104993

A few years ago it wasn’t shared between devices so each device had to do it themselves. I don’t know if it’s shared at this point.

I agree you’re not going to be training an LLM or anything. But smaller tasks limited and scope may prove a good fit.

That said, inference on apple products is a different story. There's definitely interest in inference on the edge. So far though, nearly everyone is still opting for inference in the cloud for two reasons:

1. There's a lot of extra work involved in getting ML/AI models ready for mobile inference. And this work is different for iOS vs. Android 2. You're limited on which exact device models will run the thing optimally. Most of your customers won't necessarily have that. So you need some kind of fallback. 3. You're limited on what kind of models you can actually run. You have way more flexibility running inference in the cloud.

https://www.technologyreview.com/2019/12/11/131629/apple-ai-...

But you can get an M2 Ultra with 192GB of UMA for $6k or so. It's very hard to get that much GPU memory at all, let alone at that price. Of course the GPU processing power is anemic compared to a DGX Station 100 cluster, but the mac is $143,000 less.

You want your developers to be able to do training locally and they already use Macs? Maybe an upgrade would make business sense. Even if you have beefy servers or the cloud for large jobs.

Honestly, you can train basic models just fine on M-Series Max MacBook Pros.

Do you also compare cars by looking at only the super expensive limited editions, with every single option box ticked?

I'd also point out that said 3 year old $1999 Mac Studio that I'm typing this on already runs ML models usefully, maybe 40-50% of the old 3000-series Nvidia machine it replaces, while using literally less than 10% of the power and making a tiny tiny fraction of the noise.

Oh, and it was cheaper. And not running Windows.

For training the Macs do have some interesting advantages due to the unified memory. The GPU cores have access to all of system RAM (and also the system RAM is ridiculously fast - 400GB/sec when DDR4 is barely 30GB/sec, which has a lot of little fringe benefits of it's own, part of why the Studio feels like an even more powerful machine than it actually is. It's just super snappy and responsive, even under heavy load.)

The largest consumer NVidia card has 22GB of useable RAM.

The $1999 Mac has 32GB, and for $400 more you get 64GB.

$3200 gets you 96GB, and more GPU cores. You can hit the system max of 192GB for $5500 on an Ultra, albeit it with the lessor GPU.

Even the recently announced 6000-series AI-oriented NVidia cards max out at 48GB.

My understanding is a that a lot of enthusiasts are using Macs for training because for certain things having more RAM is just enabling.

I keep hearing this meme of Mac's being a big deal in LLM training, but I have seen zero evidence of it, and I am deeply immersed in the world of LLM training, including training from scratch.

Stop trying to meme apple M chips as AI accelerators. I'll believe it when unsloth starts to support a single non-nvidia chip.

But for individual developers it’s an interesting proposition.

And a bigger question is: what if you already have (or were going to buy) a Mac? You prefer them or maybe are developing for Apple platforms.

Upping the chip or memory could easily be cheaper than getting a PC rig that’s faster for training. That may be worth it to you.

Not everyone is starting from zero or wants the fastest possible performance money can buy ignoring all other factors.

It's just more efficient to offload training to cloud Nvidia GPUs

And while the Mac Studio has a lot of memory bandwidth compared to most desktops CPUs, it isn't comparable to consumer GPUs (the 3090 has a bandwidth of ~936GBps) let alone those with HBM.

I really don't hear about anyone training on anything besides NVIDIA GPUs. There are too many useful features like mixed-precision training, and don't even get me started on software issues.

I simply want to point out that these folks don't really care about that. They want a Mac for more reasons than "performance per watt/dollar" and if it's "good enough", they'll pay that Apple tax.

Yes, yes, I know, it's frustrating and they could get better Linux + GPU goodness with an nVidia PC running Ubuntu/Arch/Debian, but macOS is painless for the average science AI/ML training person to set up and work with. There are also known enterprise OS management solutions that business folks will happily sign off on.

Also, $7000 is chump change in the land of "can I get this AI/ML dev to just get to work on my GPT model I'm using to convince some VC's to give me $25-500 million?"

tldr; they're gonna buy a Mac cause it's a Mac and they want a Mac and their business uses Mac's. No amount of "but my nVidia GPU = better" is ever going to convince them otherwise as long as there is a "sort of" reasonable price point inside Apple's ecosystem.

You can't lug around a desktop workstation.

I am honestly shocked Nvidia has been allowed to maintain their moat with cuda. It seems like AMD would have a ton to gain just spending a couple million a year to implement all the relevant ML libraries with a non-cuda back-end.

Intel seems like they could have some interesting stuff in the annoyingly named “OneAPI” suite but I ran it on my iGPU so I have no idea if it is actually good. It was easy to use, though!

https://www.technologyreview.com/2019/12/11/131629/apple-ai-...

Google did some of that on Android, which means open-sourcing their on-device TEE implementation, publishing a paper about it etc.

I happen to think they wouldn't, simply because sending this data back to Apple in any form that they could digest it is not aligned with their current privacy-first strategies. But if they make a device that still works if it stays disconnected, the neat thing is that you can just...keep it disconnected. You don't have to trust them.

It would have to be a huge conspiracy with all Apples employees. And you can easily just listen to the network and see if they do it or not.

Of course we can then worry that companies are breaking the law, but you have to draw the line somewhere… and what have they to gain anyway?

On second thought.. maybe it isn't "just like you've been doing on Nvidia cards for a while"

I can see a smaller model trained to do that may work well enough, however, I've never seen any real working examples of this work, that rabit device is heading in that direction, but it's mostly vaporware now.

They are the gatekeeper of your data for their benefit not yours.

At minimum, laws around the world prevent companies from knowingly communicating false information to consumers.

And in many countries the rules around privacy are much more stringent.

Talk is cheap and in Apple's case it's part of their PR.

Because Boeing's issues have nothing to do with privacy or security and since they are not consumer facing have no relevance to what we are talking about.

Too many holes.

You’re trusting the third party at training time, to build the model. But you’re not trusting it at inference time (or at least, you don’t have to, since you can airgap inference).

Is this the same apple whose devices do not work at all unless you register an apple account?

The advertising ID is useless as of App Tracking Transparency.

It can even be third party services sending ad candidates directly to your phone and then the on-device AI chooses which is relevant.

Privacy is a contract not the absence of a clear business opportunity. Just look at how Apple does testing internally today. They have no more respect for human privacy than any of their competitors. They just differentiate through marketing and design.

You mean .. with their said privacy-first strategy

I feel like people are being a bit naïve here. Apple's "Privacy First" strategy was a marketing spin developed in response to being dead-last in web-development/cloud computing/smart features.

Apple has had no problem changing their standards by 180 degrees and being blatantly anti-consumer whenever they have a competitive advantage to do so.

I'm not saying this to trash Apple - I think it's true of any corporation. If Apple starts losing revenue in 5 years because their LLM isn't good enough because they don't have enough data, they are still going to take it and have some reason justifying why theirs is privacy focused and everyone else is not.

Limited firsthand accounts don't really assuage my fears in the way that, say, Android or Linux does.

The future is always fungible. Anyone can break whatever trust they've built very quickly. But, like the post you are replying to, I have no qualms about supporting companies that are currently doing things in my interest and don't have any clear strategic incentive to violate that trust.

Edit: that same incentive structure would apply to NVIDIA, afaik

The logic is exactly same as ie Meta claims - we will pseudoanonymize your data, so technically your specific privacy is just yours, see nothing changed. But you are in various target groups for ads, plus we know how 'good' those anon efforts are when money are at play and corporations are only there to earn as much money as possible. Rest is PR.

And Apple's privacy stance is a moat against these other companies making money off of their customer base. So for the cost of pennies on the dollar, they protect their customer base and ward off competition. That's a pretty strong incentive.

You would be amazed at the response of some of them when I point out some shit Apple does that make their products clearly lacking for the price, the cognitive dissonance is so strong they don't know how to react in any other way than lying or pretending it doesn't matter.

I have very specific reasons to be annoyed but they are far too many to list them all in a simple post. I was working with and buying Apple stuff before the turn of the millenium and I have worked as an Apple technician and helped way more Apple users than I could care to list. This is my experience and your comment precisely illustrates that.

My comment was about the delusion of privacy first marketing bullshit that they came up with to excuse the limitations of some of their stuff.

But since Apple can't do anything wrong, we are not going anywhere. Whatever, keep on believing.

I'll check myself thanks, but you should check your allegiance to a trillion dollar corp and its bullshit marketing, that's really not useful to anyone but them.

How can they have a privacy first strategy when they operate an Ad network and have their Chinese data centers run by state controlled companies?

Apple is an amazing powerhouse but also disgustingly elitist and wasteful if not straight up vulgar in its profit motives. There's really zero idealism there despite their romantic and creative legacy.

There's always some straight idiotic limitations in their otherwise incredible machines, with no other purpose than to create planned obsolescence, "PRO" exclusivity and piles e-waste.

I mean yeah, that makes good marketing copy, but its more due to reducing latency and keeping running costs down.

but as this is mostly marketing fluff we'll need to actually see how it performs before casting judgment on how "revolutionary" it is.

> Apple's AI strategy is to put inference (and longer term even learning) on edge devices

This is pretty much everyone's strategy. Model distillation is huge because of this. This goes in line with federated learning. This goes in line with model pruning too. And parameter efficient tuning and fine tuning and prompt learning etc.

> This is completely coherent with their privacy-first strategy

Apple's marketing for their current approach is privacy-first. They are not privacy first. If they were privacy first, you would not be able to use app tracking data on their first party ad platform. They shut it off for everyone else but themselves. Apple's approach is walled garden first.

> Processing data at the edge also makes for the best possible user experience because of the complete independence of network connectivity

as long as you don't depend on graph centric problems where keeping a local copy of that graph is prohibitive. Graph problems will become more common. Not sure if this is a problem for apple though. I am just commenting in general.

> If (and that's a big if) they keep their APIs open to run any kind of AI workload on their chips

Apple does not have a good track record of this; they are quite antagonistic when it comes to this topic. Gaming on apple was dead for nearly a decade (and pretty much still is) because steve jobs did not want people gaming on macs. Apple has eased up on this, but it very much seems that if they want you to use their devices (not yours) in a certain way, then they make it expensive to do anything else.

Tbf, I don't blame apple for any of this. It is their strategy. Whether it works or not, it doesn't matter. I just found this comment really odd since it almost seemed like evangelism.

edit: weird to praise apple for on device training when it is not publicly known if they have trained any substantial model even on cloud.

The biggest players in commercial AI models at the moment - OpenAI and Google - have made absolutely no noise about pushing inference to end user devices at all. Microsoft, Adobe, other players who are going big on embedding ML models into their products, are not pushing those models to the edge, they’re investing in cloud GPU.

Where are you picking up that this is everyone’s strategy?

Read what their engineers say in public. Unless I hallucinated years of federated learning.

Also apple isn't even a player yet and everyone is discussing how they are moving stuff to the edge lol. Can't critique companies for not being on the edge yet when apple doesn't have anything out there.

I think this is being too charitable on the state of "everyone". It's everyone's goal. Apple is actively achieving that goal, with their many year strategy of in house silicon/features.

So are other companies, with their many year strategy of actually building models that accessible to the public.

yet Apple is "actively" achieving the goal without any distinct models.

You can already run those open weight models on Apple devices, on edge, with huge improvements on the newer hardware. Why is a distinct model required? Do the rumors appease these thoughts?

If others are making models, with no way to actually run them, that's not a viable "on edge" strategy, since it involves waiting for someone else to actually accomplish the goal first (as is being done by Apple).

It absolutely is. Model distillation will still be pertinent. And so will be parameter efficient tuning for edge training. I cannot emphasize more how important this is. You will need your own set of weights. If apple wants to use open weights, then sure. Ignore this. Don't seem like they want to long-term... And even if they use open weights, they will still be behind other companies have done model distillation and federated learning for years.

> Why is a distinct model required?

Ask apple's newly poached AI hires this question. Doesn't seem like you would take an answer from me.

> If others are making models, with no way to actually run them

Is this the case? People have been running distilled llamas on rPis with pretty good throughput.

I'm sorry, but we're talking about "on edge" here though. Those other companies have no flipping hardware to run it "on edge", in a "generic" way, which is the goal. Apple's strategy involves the generic.

> If apple wants to use open weights

This doesn't make sense. Apple doesn't dictate the models you can use with their hardware. You can already accelerate LLAMA with the neural engines. You can download the app right now. You can already deploy your models on edge, on their hardware. That is the success they're achieving. You cannot effectively do this on competitor hardware, with good performance, from "budget" to "Pro" lineup, which is a requirement of the goal.

> they will still be behind other companies have done model distillation and federated learning for years.

What hardware are they running it on? Are they taking advantage of Apple (or other) hardware in their strategy? Federated learning is an application of "on edge", it doesn't *enable* on edge, which is part of Apple's strategy.

> Ask apple's newly poached AI hires this question. Doesn't seem like you would take an answer from me.

Integrating AI in their apps/experience is not the same as enabling a generic "on edge", default, capability in all Apple devices (which they have been working towards for years now). This is the end goal for "on edge". You seem to be talking about OS integration, or something else.

> People have been running distilled llamas on rPis with pretty good throughput.

Yes, the fundamental limitation there being hardware performance, not the model, with that "pretty good" making the "pretty terrible" user experience. But, there's also nothing stopping anyone from running these distilled (a requirement of limited hardware) models on Apple hardware, taking advantage of Apples fully defined "on edge" strategy. ;) Again, you can run llamas on Apple silicon, accelerated, as I do.

Maybe? This is why I responded to:

> It's everyone's goal. Apple is actively achieving that goal

This is is the issue I found disagreeable. Other organizations and individual people are achieving that goal too. Google says GPT-Nano is going to device, and if the benchmarks are to be believed, if it runs at that level, their work so far is also actively achieving that goal. Meta has released multiple distilled models that people have already proven to run inference at the device level. It cannot be argued that meta is not actively achieving that goal either. They don't have to release the hardware because they went a different route. I applaud Apple for the M chips. They are super cool. People are still working on using them so Apple can realize that goal too.

So when you go to the statement that started this

> Apple's AI strategy is to put inference (and longer term even learning) on edge devices

Multiple orgs also share this. And I can't say that one particular org is super ahead of the others. And I can't elevate apple in that race because it is not clear that they are truly privacy-focused or that they will keep APIs open.

> You cannot effectively do this on competitor hardware, with good performance, from "budget" to "Pro" lineup, which is a requirement of the goal

Why do you say you cannot do this with good performance? How many tokens do you want for a device? Is 30T/s enough? You can do that on laptops running small mixtral.

> What hardware are they running it on? Are they taking advantage of Apple (or other) hardware in their strategy?

I don't know. I have nothing indicating necessarily apple or nvidia or otherwise. Do you?

> [Regarding the rest]

Sure, my point is that they definitely have an intent for bespoke models. And why I raised the point that not all computation will be feasible on edge for the time being. My point with what raised this particular line of inquiry is whether a pure edge experience truly enables the best user experience. And also why I raised the point about Apple's track record of open APIs. Which is why "actively achieving" is something that I put doubt on. And I also cast doubt on apple being privacy focused. Just emphasize tying it back to the reason I even commented.

Apple has never been privacy-first in practice. They give you the illusion of privacy but in reality it's a closed-source system and you are forced to trust Apple with your data.

They also make it a LOT harder than Android to execute your own MITM proxies to inspect what exact data is being sent about you by all of your apps including the OS itself.

The reality is, there's too much to verify, and not enough interest for the "many eyeballs make all bugs shallow" argument.

We are, all of us, forced to trust, forced to go without the genuine capacity to verify. It's not great, and the best we can do is look for incentives and try to keep those aligned.

Apple is absolutism. Even the so called "enlightened" absolutism is still bad compared to average democracy.

If you want to, for example, compile a GPS-free version of Android that appears like it has GPS but in reality just sends fake coordinates to keep apps happy thinking they got actual permissions, it's fairly straightforward to make this edit, and you own the hardware so it's within your rights to do this.

Open-source is only part of it; in terms of privacy, being able to see what all is being sent in/out of my device is is arguably more important than open source. Closed source would be fine if they allowed me to easily inject my own root certificate for this purpose. If they aren't willing to do that, including a 1-click replacement of the certificates in various third-party, certificate-pinning apps that are themselves potential privacy risks, it's a fairly easy modification to any open source system.

A screen on my wall that flashes every JSON that gets sent out of hardware that I own should be my right.

I agree; unfortunately it feels as if this ship has not only sailed, but the metaphor would have to be expanded to involve the port at well.

Is it even possible, these days, to have a functioning experience with no surprise network requests? I've tried to limit mine via an extensive hosts file list, but that did break stuff even a decade ago, and the latest version of MacOS doesn't seem to fully respect the hosts file (weirdly it partially respects it?)

> A screen on my wall that flashes every JSON that gets sent out of hardware that I own should be my right.

I remember reading a tale about someone, I think it was a court case or an audit, who wanted every IP packet to be printed out on paper. Only backed down when the volume was given in articulated lorries per hour.

I sympathise, but you're reminding me of that.

I was one of many who reported a bug in Ubuntu that went un-fixed for years, where the response smelled of nation-state influence: https://bugs.launchpad.net/ubuntu/+bug/1359836

And Log4Shell took about 8 years to notice: https://en.wikipedia.org/wiki/Log4Shell

Thanks for making my point for me.

I'm not saying closed source is a silver bullet.

I'm saying OSS also isn't a silver bullet, it doesn't find everything because there's not enough interest in doing this work.

The Log4j example alone, given it took 8 years, is enough to demonstrate that.

Everything is an illusion of trust, nothing is perfect; all we can do is try to align the interests of those working on projects with the interests of society — which is so hard that it's an entire field of study called "politics".

I would think ease of MITM and privacy are opposing concerns

If Apple enabled RCS in messages back then, but the recipient was not using RCS, then Google now has the decrypted text message, even when RCS advertises itself as E2EE. With iMessage, at least I know all of my messages are E2EE when I see a blue bubble.

Even now, RCS is available on Android if using Google Messages. Yes, it's pre-installed on all phones, but OEMs aren't required to use it as the default. It opens up more privacy concerns because now I don't know if my messages are secure. At least with the green bubbles, I can assume that anything I send is not encrypted. With RCS, I can't be certain unless I verify the messaging app the recipient is using and hope they don't replace it with something else that doesn't support RCS.

Having iMessage available to everyone regardless of their mobile OS.

Their words or your words?

By what? It's impossible for a process to know for sure if the system is rooted or not. A rooted system can present itself to a process to look like a non-rooted system if it's engineered well enough.

I'd bet that most of these apps probably just check if "su" returns a shell, in which case perhaps all that's needed is to modify the "su" executable to require "su --magic-phrase foobar" before it drops into a root shell, and returns "bash: su: not found" or whatever if called with no arguments.

That was true 20 years ago, but most smartphones these days have cryptograhically-verified boot chains and remote attestation of how the boot went.

That's just their way of walled gardening apple customers. Then they can extort devs and other companies dry without any middle-men.

Google and Samsung have been building it too, at different speeds.

Intel and AMD seem further behind (at the moment) unless the user has a strong GPU, which is especially uncommon on the most popular kind of computer: laptops.

And if you’re not one of those four companies… you probably don’t have much capable consumer edge hardware.

Watch out for being able to using ai on your local machine and those ai services using telemetry to send your data (recorded conversations, for instance) to their motherships.

This doesn’t require Ai and I am not aware of any instances of this happening today, so what exactly are we watching out for?

Now the subscription is 20$ a month and the API price is accessible. What will happen when they all decide to x100 or x1000 the price of their API ? All the companies that got rid of people in favor of AI, might have lost the knowledge as well. This is dangerous and might kill a lot of companies no ?

The big differences today that stood out to me were adopting AI as a term (they used machine learning before) and repeating the term AI everywhere they could shove it in since that’s obviously what the street wants to hear.

That’s all that was different. And I’m not surprised they emphasized it given all the weird “Apple is behind on AI“ articles that have been going around.