Hacker News Reader

Comment by quelsolaar | Hacker News Reader

Apple has fought so hard to not be fragile by trying to control its own destiny. Dont depend on Microsoft for a browser or office, dont depend on Intel, or before that Motorola for CPUs, and so on.

The problem now is that they do everything so, now they have to be best at everything.

If you want to run graphics or ML on the Mac, you have to use Apple hardware and APIS, you cant use nVidia hardware or a lot of industry standard APIs, so apple is at a disadvantage.

If you want to develop on Mac, you have to use Xcode, you cant use any other tools. So Apple is at a disadvantage.

Try connecting a variable refresh rate screen to a mac, or a VR headset, or Thundebolt device, or just a USB device in to one of their mobile devices. Try playing the latest hot AAA game everyone is talking about. Try installing a better voice assistant, store, browser or music player and you will find that Apple has made it hard or impossible.

Everything that Apple itself isn't heavily invested in and market leader at sucks. Apple isn't leveraging the industry to its advantage.

Apple is fine now, because its great at a few things and good enough at the rest of the things it needs to find a large market. How long can a company not fail at one of so many things?

View on HN

Maybe it's just me, but the integration of my Apple ecosystem makes each device feel generations ahead of my Windows/Linux devices. The whole UX flow is just so damn smooth and convenient, from AirPlay to controlling my smart home. I can't imagine doing this with non-Apple devices.

Apple does not even have to be the best at everything, or even particularly great at everything. What they need to do is be good at integrating everything, which they are.

You are right. If you do mainstream things and stay within their products it works great. The thing is that what is bleeding edge today is mainstream tomorrow. The worlds brain trust is moving away from Apple, and if Apple wants to keep up they cant rely on anyone else to help them.

That's the Nintendo problem: how do you create a platform, when your first party solutions are incredibly well-executed?

You need third parties (sort of) for scaling and long tail reasons, but it's always an uneasy relationship.

If you don't expand enough, or the market doesn't generate a solution, your platform has a gap. If you expand too much, third party developers feel burned and desert your platform.

Microsoft solved this by clearly staking its monetized territory ("the office suite"), and not really expanding in other directions (after those weird 90s attempts).

Thats a great example. I love Nintendo. Mario and Zelda is always great, but historically Nintendo has not been a good platform for 3rd party developers (I read somewhere that 30% of wii owners didn't know you could take out wii sports and replace it with another game...). The bets that Nintendo has made that hasn't been smash hits, have been terrible failures, because no one else is there to pick up the slack.

If you buy an X-Box, and the next Halo turns out to suck, you are still going to get your moneys worth of 3rd party titles.

The Nintendo comparison is missing some nuance as well. Nintendo was a gaming company that got caught in a war between the two giants Sony and MS for total control of the living room. For Sony and MS, gaming was just a single division that they were willing to take losses on or invest heavily then wait years for payback due to the consoles filling multiple strategic objectives. MS and Sony dumped tons of cash into developing their 3rd party game libraries where Nintendo relied on their in-house development to seed their consoles with games and make most of their money.

There's a big difference between the two in business model: Sony and MS had to care about attach rates and third party software sales, because they (at least initially) took a loss with each console sold: The only way they could make money was if people bought a bunch of software so they'd get licensing revenues.

The Wii was designed with different assumptions in mind, which led to a lean machine they could sell at a profit. Someone buying and Xbox and Gears of War was a huge problem for Microsoft, if someone bought a Wii and Mario Kart, it was just fine for Nintendo as long as the customer was happy: Every sale was profit, plain and simple.

I don't know about the Wii not being good for 3rd parties either, it's more that the Wii didn't do what the 3rd parties wanted to do (high tech edgelord games). So we got completely moronic things like a Castlevania fighting game, a Soul Calibur that wasn't a fighting game. How about Dead Space, a slow, ponderous horror title? Yeah, the devs turned it into a rail shooter, just about the most anti-horror genre choice imaginable.

In simple terms, they thought the Wii userbase were morons who would lap up anything - after all, they'd bought a motion control toaster to play technically really simplistic titles, and in the devs' world, technically simplistic = bad. So they gave people disrespectful shovelware, which, surprise surprise, bombed.

They had the opportunity to make solid titles - the Wii may have been a toaster, but it was still beefier than the PS2 or Gamecube which had a lot of really good titles and anyone who's played Metroid Prime knows those controls are crack.

I think the big miss with the Wii wasn't how good/bad it was, but communication. What was the philosophy behind the machine, Nintendo communicating that, the devs really trying to understand that.

> The worlds brain trust is moving away from Apple

Not really, the M1 is a smashing success with most techies. People seem to like the phones too. The watches are doing really well too. Oh, and the tablets are really popular also. Even with techies.

There seems to be a group of people sworn never to be corrupted by the evil Apple and their evil ways. I first saw it on /., and the same group has been ringing out death knells for Apple ever since.

I don't know if I live in another world but I rarely see an Apple Watch or an iPad (not anymore). I've seen maybe 5 Apple watches in my life between Europe and Asia.

I live in Europe and it's been the exact opposite. Apple watches used to seem like a novelty, but now like half the people in my social circle have one.

Maybe smart watches are not super common in general. Actually, even wristwatches are rare to see. But, among those that do wear smart watches, it's often the Apple products IME.

iPads are doing exceptionally well, I'm surprised you are unaware. They're replacing laptops for a lot of users, especially those that did not grow up around PCs. It's just that you won't often see them in public.

Apple is small potatoes outside of the USA. There is a world beyond the Valley and it's orders of magnitude larger.

I wish I had such "small potatoes" growing in my backyard ...

Apple has a significant presence in Central Europe. Not overwhelming one, American market is obviously a stronger one. It might have something to do with purchasing power parity, which is still not that good behind the former Iron Curtain. But it is a big player here, especially in the richer demographics.

I have never used any Apple products, but I am tempted. People will sing paeans about how their gadgets just work and interact together. I am more of a tinkering-hacking type, so I do not really mind delving deep into my Windows or Linux installation, but Android messes with me sometimes.

One thing about those small potatoes though: They buy software. You can actually get paid for apps on iOS way easier than Android, and it shows.

This is also true in developing Asia. I'm in Thailand and although you won't see many Thai people wearing them they all know what they are and almost all of them would like to own one if they weren't considered such an extravagant expense. Apple watch is basically a luxury item and so that means that as these other countries start to become more wealthy it's just more growth for Apple as more people start being able to afford it.

I am not saying they don't want it but I don't see it often. It is of course a luxury item, but while I see that people buy an iPhone for their social status, they probably can't afford both. And it probably comes lower in the list of social status/usage ratio as a golden watch could replace an Apple watch, but not an iPhone.

I live in Sweden, it feels like every third person I see wearing a watch is wearing an Apple watch. It is by far the single most common watch brand I see.

Same with iPads. By far the most common brand of large screen device I see. Outside of office environments I see more iPads than laptops, and even in office environments I'm seeing more iPads today than even just 3 years ago.

The Apple watch is the number one selling watch (not even just smart watch).

There's definitely a huge market of people buying it.

Whenever I go out in major cities in the US, Canada or the UK, I'll always see someone with it

I don't think people are leaving Apple hardware, but I would say they are leaving/no longer entering Apple's software stack. Native Mac dev is plummiting, native iOS dev has probably peaked by now. I see less enthusiasm jumping on new Apple features and protocols that come out every year than I used to.

It feels similar to the end of MS's domiance. They were always profitable and had huge marketshare, but just because everybody was running Windows, that didn't mean they were really embracing the full MS ecosystem. Apple feels that way today.

people have been predicting the end of Apple for 25 years. I've been there the whole time and have heard it all before.

I'm not saying its the end of Apple. There was no end of IBM or MS either, in fact while both companies were losing developer mindshare they were making record profits.

I'm talking about a much more nebulous concept of a company as a thought leader, one that people follow. There was a time when the rumor of MS entering a market would cause competition to shut down. Now nobody blinks an eye when MS ships an actual product that competes with them.

And in that regard, I think we're seeing the end of Apple's role as a technologically interesting company (at least wrt software). For example it used to be pretty much only Apple and a few exclusive MacOS developers who actually cared about design. Now designing thinking has gone mainstream all while Apple seems to be busy ignoring everything from their classic HIG.

I have quite a few apple products but none of them new. My problem with Apple is the price. My opinion is that there is always a less expensive & just as capable alternative to anything Apple has to offer, so for my use case it doesn't make sense to pay more.

The Apple products I do have were either broken and given or sold to me to repair or found at thrift shops and pawn shops for deep discounts, and they are all going up for sale soon to recoup my expenses.

That's the very reason I stay on the fringes of the Apple ecosystem. The approach of an Apple branded version of everything means that if you don't want exactly what they give you then you have a terrible time. You're either all in or all out.

I don't consider myself overly dogmatic about controlling my devices or data but the thought of going all in on one tech company for my gadgets for the next decade or so is both gross and, to be honest, really boring. I don't see how I could get excited about tech when the only choice I get is what colour it is (or the storage space).

Really boring is how I want my phone to behave: secure and boring. The fact they have a watch that works seamlessly also helps.

I don’t think of these things as gadgets anymore. In the past I’ve built a Linux DVR - so I know how to tinker and find it fun for certain things but not things I depend on and want to behave like appliances. I don’t tinker with my refrigerator or my automobile either.

I would describe modern Android as boring and not require tinkering either. Secure is tougher and up to the OEM to keep up with patches.

Certain things with iOS are just overly difficult like getting files on or off or trying to add new music to the music app without a PC.

Android is extremely secure if you don't put a bunch of relatively unknown apps on it and don't root it. I would say on par with iphone easily. I only pick iphone because they do a much better job of keeping apps and such from spying on you. Also android is constantly updating the google mothership on your location and habits.

You don’t think Apple does any surveillance and sell to companies? They don’t reject apps that spy, just the ones that do without a benefit to them.

Plus, show me another (mobile) device maker that still provides major OS updates after 5+ years.

Seriously. iOS 15 supports the 6s, released in 2015, and it’ll work fine assuming you put new batteries in it as the battery degrades.

That’s why I finally switched to iOS for my phone, and iPhones are quite affordable if you buy used a few generations old. Nevertheless, I couldn’t go back to OSX after years of using primarily Linux.

This is ultimately why I switched from android to ios. The new SE at < $200 was also a factor.

The SE is $400 for 64GB, $450 for 128GB, and $550 for 256GB.

https://www.apple.com/shop/buy-iphone/iphone-se

Yes, but you can easily find it for less than that. Here is one for $282 on Amazon: https://smile.amazon.com/dp/B088NQXD8T

I'm sure you could find even better deals on eBay.

Theres actually a deal with walmart for verizon/AT&T customers that you can get it for... $175ish I believe now? It was $200 when I did it.

EDIT: its $149 currently https://www.walmart.com/ip/Verizon-Apple-iPhone-SE-2020-64GB...

I agree this is important for security updates. However, Android is not monolothic. Browser, carrier services, even the phone app are updated independently of the operating system image for well into the future.

https://fairphone.com

Also GNU/Linux phones running mainline Linux kernel (but they don't exist for 5 years yet).

I used to think like you.. Then Apple took away my ability to fully control network traffic on my computer. Before that they took away my ability to upgrade storage and RAM. And even before that, there were always sacrifices to be made to get the Apple experience.

So yes the UX is great (I would not say amazing) and when the integration exists, the experience is almost always fantastic..

But I've always felt like it has come with trade-offs, and in recent years Apple has moved the balance enough that I no longer believe the trade-offs are worth it for me, at least when it comes to my daily driver computer.

I moved back to Windows (+Linux thanks to WSL2) in 2020 - after being a Mac-first user for almost 15 years - and am very happy with my choice.

This is exactly how I feel too...

You can imagine that controlling network traffic or upgrading RAM are not abilities that the vast majority of people are concerned about, even in tech.

Thats the thing, most people don't need those particular things. But a lot of people need something very particular that very few others need. The problem is not that apple doesn't support a particular use case, but that they cant support all niche cases.

Microsoft / Linux doesn't do everything, but if you need something specific you can bet that there is someone out there that provides that functionality for those platforms.

They're concerned about it, but non-technical people get hooked on the familiar. They often don't want to learn a new UI when they know one already.

I have a friend who has a 2015-ish Macbook with 4GB RAM - and she is getting into Photoshop.

So her computer works well for her. The monitor is good enough, the CPU is good, and so on. 4GB just won't cut it.

Since she can't upgrade her RAM for $150 like she should, she has to get a new computer. But she likes the trackpad and the Apple UI so much, she is going to get an iMac with 16GB RAM and 256GB SSD for $1600, instead of a Windows PC with a top-line processor, 32GB RAM, 27" screen and 1TB+ of SSD.

If you want to do mouse-intensive work on a laptop the mac trackpad is the only way to go.

The retina screen alone would cost $400+ for a windows. And the Mac is much more of a pro tool when doing creative work. My windows machines always crash on me.

I haven't had windows crash on me when I wasn't doing driver stuff in literally years. Something is wrong with your hardware, unless you're talking about the software that you are using on windows. If those crash it's not windows fault it's just bad programming.

If that’s the case you are the lucky one. It’s surely not a coincidence that Windows software suffers from “bad programming” 50x more often.

There is simply no contest in reliability, you’re always expecting a new problem with windows - just yesterday, in fact, an OS update disabled my network card and I had to reinstall everything to “fix” it. You can use a Mac for years without even knowing what a driver is.

I use both macOS and Windows daily, and I think I saw like one BSOD in the last year, and quite a lot more kernel panics on macOS.

Regarding the software I use, I wouldn't say that it's any more stable on macOS than on Windows.

Working in a company where I deal with equal amount of Macs an Windows, we spend a lot more time dealing with Mac issues that simply cannot be solved due to either hardware or software limitations Apple have put in place.

Windows is very solid and we rarely see issues with it on decent hardware which still costs 1/2 of what a lower spec Mac equivalent.

Of course it's not a concern for all users, definitely the network one, but I think you're underestimating how commonplace RAM upgrades are.

In my own circles (so anecdotally) I've had plenty of conversations over the last 20+ years about helping tech and non-tech people upgrade RAM on their computers, including laptops.. And in the last 5 years I've heard plenty of people be surprised (and bummed) that they can't upgrade the RAM on their Macs post-purchase anymore.

And of course maxing out the RAM at purchase time is a complete rip-off compared to market prices.

Every Mac I've owned (5 machines across desktops/laptops from 2006 through to 2016 I think), I eventually upgraded the RAM myself via aftermarket parts at reasonable market rates and it greatly extended the life of the computer.

Given the premium cost you already pay for a Mac, not being able to do this anymore is a deal breaker, for me at least.

I spent several minutes trying to get my AirPods to connect to my phone today instead of my iPad. Ended up being late for a call because of it. Happens a few times a week. More than irritating. First world problem, to be sure. But still the whole “it just works” line irks me no end.

Sounds like a problem with your AirPods or your iPad's saved Bluetooth device settings.

I've literally never had a problem for years. Switching is seamless. I don't even have to manually switch, it happens automagically. Same for everyone I know with AirPods.

> from AirPlay to controlling my smart home. I can't imagine doing this with non-Apple devices

That's because many smart home devices are only supporting the duopoly OOB due to their market prominence, indeed both Apple & Google are doing something good enough to retain their *poly position.

If the smart home devices have open API, then there's no reason for it to work great with anything else. A case in point, put KDEConnect on Android, Linux Phone or Sailfish and you can even play the audio from your desktop browser on your phone speaker.

>then there's no reason for it to *not work great with anything else

> from AirPlay

It’s funny that you take AirPlay as an example. I would have agreed with most other things, but AirPlay is light years behind stuff like Google Cast or Spotify Connect.

It ain't. It's also far easier to integrate everything when you control everything. My current Samsung tablet and phone have the same levels of integration as my iPad/iPhone did. They can share calls, texts, pass files etc. Windows does a good job integrating with my phone too.

Smarthome is one place Apple fell down. I tried to add my hue lights and instead of just automatically finding the hub like Amazon and Google I was supposed to take a picture of a sticker? Could never get it to work. They also took far too long getting an affordable voice speaker for the home out. I'm not digging my phone out every time either.

It all works beautifully until it doesn't then it all becomes close to worthless.

Take getting videos off an iPhone for instance, something I have to do many times a week in my job.

If you have AirDrop it's effortless. But my MBP has a faulty bluetooth chip so bluetooth generally stops working after the machine has been running a few hours so I need to reboot every time I move videos from my phone to my machine.

Are there alternatives? Nope not really, moving iPhone videos to a PC, Which most video editing workstations are now because Apple neglected or fumbled the high end for close to a decade. You can't AirDrop so what do you do?

- WhatsApp? Nope has a 60 meg limit - Slack? Nope, for some reason trying to send video files 90% of the time stalls or just takes close to 20 minutes. I think they have some built in throttling to prevent my using their service as a AirDrop-drop in because airdropping then sending the same video from desktop is fast - Email? Nope, files are too big

Only real solutions are AirDropping to a Mac then sending on slack or over network or trying to figure out how to do it over the network via the Files app which is also slow and weird.

All that beautiful interoperability goes from being a benefit to actually making the product look bad the second it can't just interoperate with things you need to get your work done.

End of the day they can't do everything and are unwilling to create machines that are competitive in these niche areas but once you're in this niche the whole Apple philosophy starts to look like a bad one.

All it would take would be for them to release an AirDrop for Windows app which would make their platform look better, I'm still your customer even when I have to send a file to a Windows machine Apple... not providing that feature isn't making Windows look bad because the reason I'm there in the first place is because it outperforms even a $20K Mac Pro.

Couldn’t you just use Dropbox or iCloud Photo Library? Yes you would have to pay.

I think you can transfer photos and videos for free using a usb cable as well. https://support.microsoft.com/en-us/windows/import-photos-an...

These came to mind right away and work for both macOS and Windows:

- Upload to a cloud storage service of your choice and download on the other end—Google Drive is very fast

- Use iCloud Photos (there’s client software for Windows too)

- Plug a cable in and drag and drop like it’s any other digital camera

- Create a Windows SMB network share (or Samba on other platforms), connect to it from Files, and use the share menu from Photos to put it in the network share

> Plug a cable in and drag and drop like it’s any other digital camera

Yeah, this rarely works. It takes a long time to acknowledge there's any files, and then it tries to copy them, and then crashes instead. Over and over. Turning off the mechanism that autoconverts image files from HEIC to JPG helps (I never tried JUST video files, but those are unreliable either way in my experience). Obnoxious, overall.

What about copying to a USB memory?

Straight from the phone? How do you do that?

Hmm I was thinking about a USB cable to the laptop (I do with my Android), but wrote USB memory instead.

Now I see in the comment above that cables often don't work.

I wonder if there're apps for saving files to USB memories

Tip: You can connect your phone to you macbook and use build-in preview.app "File -> "import from YOUR_IPHONE_NAME" (however since the last few macos releases preview.app crashes like 50% of time when I'm importing). This only works for movies/photos though and only when using macos.

Although I agree with general sentiment. I used to have android phone and I really enjoyed that you could just connect your phone with cable and browser data with any file manager.

I always have a bad mood when I have to download/upload/sync something via itunes, e.g. voice recordings / pdf / documents / ebooks

In my experience - AirDrop works as expected about 30% of the time. Enough to make it absolutely by far the worst option for file sharing.

I don’t even consider it an ‘option’ anymore while transferring files. I’ll just upload it to my server, since I know - by default - even on the same wifi network, AirDrop is as useful as picking a number in your head and having a friend guess it.

I can’t honestly think of a regular ‘service’ on iOS and MacOS that is as fundamentally broken as AirDrop. I’m not sure what’s so complicated about writing a device to device file sharing service, but - for instance - I just tried to share a file from my Mac to my iPhone. Didn’t work; don’t see my iPhone. Same wifi network - no firewall; no filters. This is a majority of the time.

Yeah, AirDrop is great when it works. However, when it doesn't work (happens sometimes to me), it cannot be easily fixed.

Sounds like you need a new MBP?

You can use "image capture" app on mac. Also if you want you can use the ifuse method. https://github.com/libimobiledevice/ifuse

But yes, Apple should have made it easier.

I'm sorry, but this is a really bad take. The premise of the argument, namely that Apple does (or attempts to do) everything, and does not leverage the industry to its advantage, does not stand up to scrutiny.

Apple relies heavily on a modularised supply chain. It designs its own chips, true, but relies on TSMC and Samsung to fab them. It designs its own camera system, yes, but buys the crucial components (the sensors) from Sony. It designs its own screens, but uses Samsung technology and production capability to do so. It runs its cloud services on top of AWS and MS Azure. And famously of course, they outsource production of almost all their devices to Foxconn.

Apple carefully selects all of this and makes it work; the "take", as it were, wasn't precisely that Apple does everything (and so I feel like you critically misunderstood it), but that Apple controls everything: any extension of the ecosystem doesn't just block on Apple... it nigh unto must be initiated by Apple.

This. Plus, their dependence on Chinas goodwill makes them vulnerable.

China commands significant power over Apples manufacturing capabilities. Just take a look at Apples supplier list ^1 This is the opposite of antifragile.

Tencent, Pinduoduo, Alibaba are prime examples of how much influence the Chinese government is willing to exercise over companies in case it finds them to not be aligned with it's own interests.

The Chinese government effectively has access to the private iCloud keys for Apples Chinese users.

From an investors perspective I'm a bit bearish on AAPL.

^1: https://www.apple.com/supplier-responsibility/pdf/Apple-Supp...

TSMC isn't going anywhere and semiconductor production for 3rd parties at Samsung isn't going anywhere either, they do tons of it. I know lots of people that work there. EVERYONE (almost) outsources semiconductor fabrication. Are they all fragile? Not many companies own their own plastic or metal factories either. I really am not sure what you're trying to get at here.

>If you want to develop on Mac, you have to use Xcode, you cant use any other tools. So Apple is at a disadvantage.

I don't think that's true even if you consider IDEs that target the OS X frameworks only. Intellij has a product, Flutter targets iOS, I am not an OS X developer but those two jump out straight away.

As for developing on Mac - it's a Unix & so a very capable development tool to target a range of platforms.

Even though you can develop in Flutter, you still need xcode tools to build the apps, even if you can mostly skip the IDE.

We should differentiate between Xcode the ide and Xcode tools, which are just git, clang, etc. The tools aren’t apple’s product.

In Linux they would have been separate packages, but I don’t think the App Store has dependencies so how else are they going to get the required packages installed for their ide.

In this case the argument is focused on Apple forcing you into their ecosystem. Even though some packages might be generic, Xcode tools & codesign wont run and build on another OS then MacOS.

This is just like saying clang compiled for Linux won’t run on Windows [1].

[1]: Yes, I know about WSL. No, that doesn’t count.

It's not: for any other OS, you can develop and compile on whichever other OS you want. Sure, MSVC won't run on Linux, but you can just compile it with clang or GCC on Linux and target Windows. There is no way to do that for iOS.

And this is not because no one wants to write a backend to target iOS, it's because Apple won't allow it. If you want to run native code on iOS, it must be built and signed with XCode on a Mac.

Try to create a UWP application or Windows service, or do any kind of debugging or ETW monitoring, like that.

Fair enough with UWP. Might be one of the many reasons no one uses it...

I will still note that debugging and monitoring are usually not going to work cross-platform - try to log to journalfs from a Windows system (wihtout using MinGW or similar).

Plenty of people use it, they don't need help from Linux devs trying to cross compile applications without any tooling for UI/UX.

You only need clang and codesign to build an app.

If it were actually this simple in practice I wouldn't be jumping between AppCode and Xcode all the time

Could you provide a few examples from your workflow that requires you to open Xcode?

i think people might be conflating needing to install xcode vs launching xcode

if you want to build ios apps installing xcode is a requirement (to get the proprietary cli tools) but launching xcode is not required

to build native mac apps, neither installing nor running xcode is necessary, as you can build with the smaller cli-only tools which include only mac os headers/libs

If you want to do real-world things, like change your project settings across different compilation targets (AppCode has a buggy/incomplete settings interface that can't be trusted), use the interface builder (inc. previewing your UI, inc. creating/editing the UI if you're not using SwiftUI), debugging the UI hierarchy, performance profiling, etc., you'll be switching over to Xcode. Once again, in theory you could just have everything running off whatever `make` toolchain you prefer, but if you're making apps for a company, and sharing code with other team members, chances are your exotic setup isn't going to cut it.

Anyone that has written portable UNIX software knows why stuff like autoconf even exists, targeting POSIX is just like trying to make HTML/CSS/JS, it works as long as everyone uses Chrome, sorry GNU/Linux.

As for those IDEs, they are a subpar experience of what the whole AppFrameworks are capable of, just like Borland of yore always playing catch up with Windows SDKs.

> Targeting POSIX is just like trying to make HTML/CSS/JS, it works as long as everyone uses Chrome, sorry GNU/Linux.

That's a bad analogy IMO. GNU/Linux is mostly POSIX compliant. There are some edge cases one runs into with C programming, but if talking about shell scripting, sticking to POSIX is definitely more portable. Bash would be a chrome equivalent, otherwise it's like targeting a subset of what every browser can do. Testing against Busybox/musl (Alpine) is a good way to assess portability.

Try to fork in a POSIX compliant way an application using signals and threads, across several UNIX implementations.

Or asynchronous network programming with certificates, while using only POSIX APIs.

Dealing with filesystem capabilities, file locations, daemon configurations, port management utilities,...

Ah, and anything GUI related is not even part of POSIX.

As a FreeBSD user, I don't think I ever had a C program that didn't advertise being tested on FreeBSD work first time. There's no test suite to verify that you've used only the POSIX subset of what Linux provides, and it's not something you're going to get right solely by trying, no matter how careful a programmer you are.

Developing good software for Apple's platforms without Xcode remains an unsolved problem.

As was mentioned here, xcode is not a requirement. I manage my development tools on Big Sur with NetBSD's pkgsrc via pkgin.

Not strictly required, but practically implied.

Most of this are factually untrue. Do you use mac? Like you can play music with VLC and use VS Code or soma JetBrains IDE. Most developers who use mac never launched Xcode, because, you know, it is mainly to make iOS/macOS apps, and it by far the best tool I seen to make mobile apps.

Do you often run ML jobs on workstations? Why? Yes, macs are not great for gaming, oh well.

> it is mainly to make iOS/macOS apps

You are talking about developers using a Mac, who also develop for the Mac, as if its this tiny corner case that isn't important. Maybe thats true. But it should be really important for Apple that those developers target the Mac, thats how Apple gets 3rd party software.

My intel MacBook Pro runs AAA games just fine on it’s Windows Bootcamp partition.

> Try connecting a VR headset to a Mac

Well, try connecting a VR headset to anything that isn't a Windows PC.

> Try playing the latest hot AAA game everyone is talking about.

Again, this is mainly because of the dominance of Windows, and also to some extent DirectX.

My personal take is that, when left to their own devices, developers and the industry will focus on the dominant platform first (Windows) and other vendors (Mac) will tend to be an afterthought. This was more true 10-20 years ago, when Mac wasn't as dominant.

Because of that, in order for Apple to be successful, Apple couldn't rely on the industry to make their platform great, so they had to build their own 'exclusives'.

Taken another way - in a parallel universe where Apple hadn't heavily invested in building a lot of this stuff themselves and relied on leveraging the industry, I'm not convinced that they would be alive today.

> Well, try connecting a VR headset to anything that isn't a Windows PC.

I have never tried a VR headset under anything that wasn't Linux, so our experiences differ on that point.

I'll concede the point that SteamVR on Linux is possible because of Valve after some research, but I still think it's an example of were Mac can't rely on outside industry because they will only support the dominant platforms or their own platforms (for instance, Valve supports linux primarily because of steamboxes).

And Valve not supporting Mac again shows that relying on outside industry is not a viable strategy if you are not the dominant platform. Mac wasn't a big platform for them to sell VR titles as it's not used by many consumers to play VR titles, so they drop support.

As another example, the Oculus client only supports Windows (outside of the Quest etc). There are mac's out there with powerful graphics cards and lots of RAM, however they are few and far between compared to Windows boxes, so Oculus chooses Windows as the dominant platform - If you are going to support one, it will be Windows, and if you are going to support Mac, well it's not going to be the first-class citizen. If Apple wants VR support they are going to have to go out there and graft for it.

The overall point is just - Apple is successful because they decided they wouldn't wait for industry to support their platforms, and that they would just build it themselves if the industry didn't come. For instance, iWork is a direct result of Microsoft not bringing Office to Mac originally - The lack of good presentation software was going to be something that prevented professionals using Mac, so they went out and built Keynote themselves. On the flip side, Photoshop was a great app with best support on the Mac, so they never needed to build a good photo editor. The investments are all to fill in gaps where other people in the industry weren't bringing stuff to the party.

> And Valve not supporting Mac again shows that relying on outside industry is not a viable strategy if you are not the dominant platform. Mac wasn't a big platform for them to sell VR titles as it's not used by many consumers to play VR titles, so they drop support.

Yes and no - I think part of that was also that Apple is openly hostile to other app stores on the Mac. It's true that third parties are lukewarm in their support for Mac, but it's very hard to disentangle to what extent that's the effect of Apple's lukewarm support for third parties rather than the cause of it.

The reason VR headsets don't work on Mac isn't because Valve decided to not support them... VR headsets typically plug into the HDMI port of "the latest greatest graphics card", which isn't really "a thing" on Apple computers (until maybe extremely recently--like less than two years ago--on the Mac Pro, and I am not even sure about that), and before you say it is because nVIDIA is simply not bothering to support Mac, I'm pretty sure it is because Apple doesn't really provide sufficient hardware ports or driver extension points for them (and then probably will get pissy about how they have lost control over color balance or something as their software probably only works "perfectly" with their hardware, etc.)

Apple deciding to drop support for industry-standard cross-manufacturer APIs sure didn't help.

Unless you specifically need to target Apple, you don't come across their ecosystem of APIs, which makes it harder to maintain cross-platform compatibility.

Reportedly, Apple is "all-in" on xR, so wait and see I guess. But they're not on the public supporter list of OpenXR for now, as a datapoint: https://www.khronos.org/OpenXR/

Also, Apple may have a vendetta against nVidia. It's been ages since any nvidia product worked on macos.

>For instance, iWork is a direct result of Microsoft not bringing Office to Mac originally

When was that? MS Word has been available on the Mac since 1984, Excel since 1985 and PowerPoint since 1987.

https://en.wikipedia.org/wiki/Microsoft_Office#Mac_versions

Microsoft released Office for Mac before there was even a Windows version. Over 10 years before Keynote even launched. Office v. X released the same year as OS X, two years before Keynote and 4 years before Pages.

Keynote was made because Steve Jobs wanted better presentation software than what others provided that was more like Concurrence on NeXTSTEP.

Tech has this perversive notion that you have to be either the objectively best possible option or you’ll die. And this opinion seems to be held by the same people who have been proclaiming that Apple are going out of business because they can’t compete with Linux or Lenovo on pricing and that their mobile phone would fail for lack of physical keyboard or that they where a joke as a mobile company because “they only sell one model of phones”.

You don’t have to be the best, you just have to be your customer bases favorite choice. And Apple is great at delivering to and making money off their customer segment. No need to have a monopoly, no need to “slay the competition” they are just doing their thing making tons of money on a smaller segment of the total market. Imagine if people assumed the same about car brands, that there was only room in the world for a single car company. Sounds absolutely absurd doesn’t it?

Perhaps we can entertain the thought that Apple might not be going away even though they’ve been going out of business since 1976 according to public opinion.

It may be worth watching D5 again, but here is the bit [1]

>We wont so good at partnering with people, where as Bill and Microsoft were really good at it. They learn how to partner with people really well. And I think if Apple had a little more of that in its DNA, it would have become more successful. And I dont think Apple learned that until few decades later. - Steve Jobs.

I think Steve Jobs was way way better with partnering in his return from NeXT. From iPhone with Cingular to patent signing with Qualcomm partnership. He really knew which battle to fight and which to work with. Tim Cook seems to have a very bean counting view on almost everything. To the point of not valuing much partnership. Which may be a good thing in Supply Chain but not so good with other businesses.

[1] https://youtu.be/wvhW8cp15tk?t=5012

Nvidia wants you to use CUDA instead of OpenCL..

Thankfully, otherwise we would be stuck with compiling C on the fly without any kind of debugging experience other than GPU printfs.

Your argument is basically what everyone said in the introduction of the article and in fact what was generally said even before the iPhone came out. It's not even a new situation for Apple to be in. 20 years ago it was companies willfully not wanting to support apple for their game or software or whatever. I think companies have come a long way on that.

So they are not commoditizing their complement [1] in general (app store excepted!)?

[1] https://www.gwern.net/Complement

What are you talking about? The literally outsource all their manufacturing.

They didn’t even write their own OS or browser or compiler. Microsoft at least does that.

> The problem now is that they do everything so, now they have to be best at everything.

They certainly have the cash on hand to actually do so.

> If you want to develop on Mac, you have to use Xcode, you cant use any other tools.

Unlike Windows you can do everything you want in free Xcode, only when you want to distribute your applications to others you need the 99$/year codesign certificate. Also, you don't even need full Xcode to compile programs, see e.g. Homebrew or MacPorts.

> Try connecting a variable refresh rate screen to a mac, or a VR headset, or Thundebolt device, or just a USB device in to one of their mobile devices. Try playing the latest hot AAA game everyone is talking about.

Most of what you mentioned is gaming, that has never been any market of Apple's - and besides, you can always dual-boot into Windows on the Intel side (and probably, soon enough into Linux on the M1 side). The other stuff (TB, USB on mobile) isn't well supported on the competitions' side too - all I can plug in into my Android phone and tablet is a mouse, keyboard and USB soundcard.

> Try installing a better voice assistant, store, browser or music player and you will find that Apple has made it hard or impossible.

Agreed on the voice assistant. The rest? I can download and install apps like I want, I am not depending on the App Store for anything. Browser? Chrome or Firefox (although I admit, I would be happier if Apple opened up and documented the ways that Safari can use to save battery life) can be downloaded and installed in minutes. Music player? Spotify and VLC have all you want, Amazon Music has an OS X app, and for the iOS side foobar2000 also exists.

> Apple is fine now, because its great at a few things and good enough at the rest of the things it needs to find a large market. How long can a company not fail at one of so many things?

Apple has enough cash on hand and cash flow from their current product line to wait out any "failure" - and they have the cash to buy so much advertising to make a literal pile of poo a bestseller anyway. They are so huge they literally cannot fail.

> Unlike Windows you can do everything you want in free Xcode, only when you want to distribute your applications to others you need the 99$/year codesign certificate

Are you comparing this to e.g Visual Studio now? Because most SDKs are free (.NET, even the C++ compiler can be used freely now afaik), as is vscode and there are of course several other free IDEs.

I was referring to professional usage of said IDEs, comparing the vendor-provided solutions. Unlike Xcode, as a business above six users you will need one of the commercial licenses for VS.

vscode is vendor provided, free (also for enterprises), and can be used to build all kinds of apps including C# and C++ (using msvc, clang, mingw etc) It’s true that VS is not free for commercial use yet.

Not quite true, there are some subtleties on your argument.

VS Code supports a subset of Windows development workloads, try to develop C++/WinRT applications or a WPF GUI in VS Code, for example.

VS Community is only free for open source, students or companies up to 5 employees with a specific yearly revenue limit.

There are certainly features in VS that are not in VS code (otherwise the install size of what 20x? would be pretty crazy). Tools like profilers for example. Visual designers are also among them but are not a requirement. I develop both WPF and WinForms and don’t use designers even though I have had VS enterprise for the last 20 years.. so the question is “can you do it” and “do they have the same feature set” which are obviously two separate questions. There are features available in vscode that aren’t available in VS too.

FWIW, I have been doing Windows development on and off since the 90s, and not only have their free tools have always been extremely capable, you can get third party and open sources tooling for literally everything... with Chrome having finally ditched msvc a few years ago for clang, I feel like there is really no excuse anymore for feeling stuck with Microsoft's tooling.

The issue with macOS development is that there are simple things you pretty much have to do for even the most basic app to work correctly--such as build the menu bar from a nib or bundle icons--that require you to use closed source software like altool and ibtool that come with Xcode (and then, despite being command line tools, for some incompetent reason tend to be tied very tightly to the current version of the operating system, which makes stuff like CI extremely annoying).

Yeah, when one likes to only do low level C and C++ Win32 code, with crutches for everything else.

I mean, I feel like you are treating something I consider "amateur hour development" as "the only reasonable way to do things"; but, even so: you seem to be missing that on macOS it is nigh into impossible to opt out of sitting on a Mac using sad software that comes with Xcode, whereas even C++/WinRT (which you mention, but which I would still mostly consider "low level") is something I can easily use with an alternative IDE from a different operating system. The only place where I feel like you will be sad with Windows development sitting on a drastically different stack is dealing with the more cutting edge .NET functionality. On macOS, it is like pulling teeth to get even a hello world app working without Xcode. Alternative stack tooling like Flutter from giant companies like Google still is forced through Xcode workflows.

I would like to see you use Visual Studio plugins (C++/WinRT is one), or do ETW monitoring, or Event Viewer plugins from Linux.

So many examples that are beyond bare bones Win32 C calls.

Which is reasonable. If you can afford 5 fulltime employees you can afford visual studio or take the plunge and move fully to vscode.

> Apple isn't leveraging the industry to its advantage.

Apple is the industry. When you're the industry, you don't leverage it, everybody else attempts to leverage you.

If for some reason they find themselves needing to eg switch back to CPUs supplied by someone else because they can no longer be great at that - given their scale, Intel or AMD (or whomever) will get on their knees and crawl begging for the business.

None of what you're claiming are disadvantages are actual disadvantages unless they lose the iPhone's position. Unless that happens, they are the industry, they are the (increasingly) dominant ecosystem, the advantage is theirs to press. They're winning.

Reply on news.ycombinator.com