The Mac price crash of 2021
zdnet.com
zdnet.com
I am going to think real long and hard when I buy my next laptop.
In fact, I reckon you'd also ask how the currencies are staked against each other. All that. Won't you?
Can you boot from external drive?
Previous company required me to leave work laptop in office that's why I bought one. Of late this company has been pressuring to install surveillance apps so my new-found interest in reviving my old laptop personal usage - browsing, mail etc - I don't anything else from my personal laptop.
So essentially it has not been used since 2017 (I keep it safe in an airtight bag). It boots up and shows the [?] as in no disk found.
In the US market there are some power sellers who say 2 year warranty and get big prices. Meanwhile while I was selling 1 as a non-prof seller I got tons of questions of "can I return it" and "what is the warranty" and two times my item was bid up to the stratosphere by non paying biders. Finally I listed it buy it now only and eventually a no fuss buyer bought it for half the price of the prof-sellers.
My point is, ebay prices don't always tell the whole story.
Here's the £850 one - does have a TB SSD but still...
https://www.ebay.co.uk/itm/apple-macbook-pro-retina-15-4-Inc...
And the £620 with 256GB https://www.ebay.co.uk/itm/Apple-Macbook-Pro-15-inch-mid-201...
That is some serious value being retained right there.
Saying that, I still own a MacBook Pro 2013 and a newer 2016 for from work, and the performance is indistinguishable - so maybe £750 is a good deal.
That is all part of the reason prices stayed high - until there was a genuine update.
It's called a dead-cat bounce.
I thought the main strengths distinguishing the performance of the M1 were the idle power chips that saved a lot of energy, and the shared CPU/GPU memory. For loads running flat out all the time and emphasizing GPU, the difference is not as great, I thought?
Apple is still selling Macbooks with Intel inside. That cannot be just stupidity, right?
No, it just means the transition to ARM is not done. At WWDC they said it’ll take two years to convert the whole product line [1]. We are only a few months into that. Judging by the last couple years, the Apple HW ship turns slowly, but it does turn.
Also, there’s no need to be so harsh. When other people do something you don’t understand, it’s bad to assume they’re stupid. They probably have reasons you don’t know about.
[1]: https://arstechnica.com/gadgets/2020/06/this-is-apples-roadm...
Look up the performance stats, for loads running flat out or using GPU, M1 is insanely faster. People are replacing their two year old $5K top of the line MacBooks with an entry level M1 and seeing massive real world performance gains. That’s the point of this article - why would anyone pay $2-3K for a used 15” MacBook Pro that retailed for $5K in 2018 when they can buy an M1 that blows it out of the water brand new for half the price?
There are still limiting factors (if you need ports, more RAM, etc) since these are entry level models, but the rumors about the next gen MBPs coming in H2 of this year sound like they’ll tackle those easily.
I don't think it's worth struggling with a slow computer until the end of the year if you'll eventually buy a "spec bump" of the current two-port M1 macbooks.
The first Apple Silicon release is better than nearly everything available from Intel. That same first release will likely be obsolete by the end of 2021, as Apple advances with M1X or M2. That annual obsolescence may well continue for the next ten years, too.
So the used Apple laptop market is going to look a lot more like the used iPhone market - and in the used iPhone market, you get 15% of what you paid, not 85%, because there’s an endless supply of new phones forcing the used market down.
Post-pandemic, I’m almost exclusively typing on a Kinesis, but pre-pandemic, I found the laptop keyboard to be okay.
If I’d known we’d have a pandemic, wouldn’t be traveling in 2020, and that M1s were coming out, I’d have waited longer, but I’d probably have made other, more impactful decisions than which laptop to buy.
If you can find a much newer one, the more recent 16” macbooks have a much improved typing experience over the first attempt at the butterfly keys, otherwise a 2015 one with upgrades to at least the RAM will still run quite well.
I'd be surprised if they brought back ALL of the ports. I'm guessing they'll bring back the HDMI and the sd card slot (although probably microSD this time).
As such they would need to add a full SD slot for people who use professional cameras because that is the form factor that these cameras use -- not microSD.
Think about it another way -- what kind of products use microSD cards? What market would Apple be targeting if they brought microSD slots to their laptops?
I don't use mine for CPU intensive tasks though, so YMMV.
Wife's Macbook Air needed a new battery end of last year and I thought "eh, $129 is worth it" only to find out that the battery wasn't actually available at any price from Apple. Ended up replacing it with an aftermarket $30 one and it's been great for 3 months, but that's also not a glued-in battery, so it was an easy 10 minute job, certainly not the case on the 2015 MBP.
If your computer is coming up on one of those timelines, the battery is slightly weak, and you want to keep it, might be worth getting the service while it's available. I was not happy to have to go aftermarket, given some of the shadiness in that market.
People want computers perhaps more than normal, but do they have the money to spend?
I think this is why Apple was advertising Education pricing on the M1 Macs the day they came out. They don’t care that you qualify, it’s basically a “buy from Apple directly instead of a third party retailer” discount.
This is all just my amateur theory - I think Apple wants to keep volumes up and sell something to meet demand.
When things turn around, I'd expect Apple to pick back up pretty well. I'm not an Apple owner nor a fan, but I definitely have an interest in watching the evolution of these since they're the first real rival to x86 in a long time.
There's a real possibility I'll end up going to a future Apple, Microsoft (since they're looking into doing their own ARM chips now as well), or some other ARM architecture in the next 10 years if Intel and AMD can't keep x86 ahead. It looks like Intel already lost with almost no hope of getting back into a competitive position, so AMD's what I'm watching closest on that side.
Also, look, before anyone downvotes me for being Anti-Intel(tm), I just want to say that I'm going to remain skeptical until Intel can do 2-3 generations of good chips in succession without dropping the ball hard. Zen has two or three generations of good improvement showing up, and Intel might have ONE generation with the newest chips coming out but that doesn't mean they'll keep it up.
If Intel does that, I'll keep them in consideration. Until that point.. well, I'm just going to remain skeptical and cautious.
I've heard nothing but complaints about their licensing, but IIRC some older versions (CS6?) still work after only one-time activation.
I'm usually weary of GEN1 products, but it's been hard for me to hold off on picking up one of these. The performance boosts seem to be quite impactful in a daily workflow especially if you're compiling a lot of code regularly.
I'd be upgrading from a late 2015, 15" non-dedicated GPU macbook pro. It works fine and I don't really have any complains other than of course I'd love if it were faster. Since I'm on my computer all the time, removing sources of friction are important.
With the change to ARM you'll now build an ARM container (or exe, etc.) and it won't run on Intel or AMD-based servers. You'll need to invest time figuring out Rosetta 2, cross-compilation, QEMU, and other multi-architecture workarounds to produce non-native binaries. It's not difficult or particularly confusing once you understand all the moving parts, but it has tripped a lot of people up who thought things would just magically work.
If you're dealing with particularly gnarly messes of complex build systems with obfuscated and obtuse decades old makefile messes it can be quite daunting to add cross compilation and multiarch support. But most codesbases aren't like that and you'll probably be fine. Interpreted and JIT-compiled languages like Python, Node, etc. are generally smooth sailing (unless you link to native code). Modern systems languages like Go, Rust, etc. are very easy to cross compile. With C/C++ you're up against a bit more pain depending on how nice your build system behaves (but let's be honest if you're dealing with a C/C++ codebase in 2021 you're used to some pain).
The inverse problem is also now true. If you pull down a container image, download a binary, etc. it ideally has to be ARM to work. Rosetta 2 can help in the short term to get Intel dependencies working, but again there will need to be some work done to get the dependency moved to multiarch and ARM support.
There's also a major gap in this story if anyone on your team is using Intel machines. There is no good equivalent to Rosetta 2 for Intel--i.e. those Intel users are going to have trouble making an ARM binary or container that your mac can consume. You're going to have to coach them up and walk them through using something like QEMU to achieve this goal (this might not be easy, be warned). IMHO it will be easier to push this load onto your CI system (assuming it can support building ARM) so that people just push source code from their machines and CI takes care of ensuring Intel, ARM, etc. builds are available to everyone.
If this sounds like a lot of issues, try to stay on the happy path of using one architecture everywhere. M1 macs + AWS Graviton 2 instances in AWS is 100% ARM64 all the way and really something incredible once you live in that world. Other cloud providers (ahem Azure?? Google??) need to get their act together and get ARM64 instances available yesterday.
For context, I'm a dev. I do DotNet Core, Node, Go, Python, and Ruby. All of them are far faster than previously on the Intel Air, and at least as fast as the L14 with its 16GB and Ryzen 7 Pro. The only performance downside being an initial hit of up to a second or two for the tooling the first time it's used since the last restart (whilst it does its Rosetta wonders). Everything is working as it should, although I have no current need for Docker so can't comment on the progress of that. I had some issues getting Homebrew using ARM versions of stuff (I think, it was a while ago now) but otherwise life is good.
The biggest benefit for me, though, is that unlike previously I am getting this enhanced performance from a stone-cold fanless machine with a battery life good enough that I only plug it in daily out of habit (I could easily go several days without).
Two ports is enough (for me) - bear in mind that really is two ports, as mostly there is no need to be connected to the power so you don't lose one during the day.
The keyboard isn't a patch on the ThinkPads, obviously, but its the proper Apple standard not the dodgy one from a few years back. Feels good, slightly shallow travel. One oddity is there are no longer hotkeys to control key backlights. Unless you open the settings, MacOS does it automatically. I didn't think I'd like that, but it works consistently well.
BTW in the couple of months I've had it, I've had three lock-ups. All when running JetBrains IDEs as it happens, but that's Java for you. Rebooting is quick enough that it wasn't an issue.
Everybody has hated on Microsoft my whole life, but you know, I can't remember the last time Windows 10 locked up on me. I don't even know if it "blue screens" the way windows used to. For all its faults, it "just works".
When I run Windows 10 Professional at home (which I still occasionally do) I get excellent reliability. The moment it hits heavily locked down networked office laptops is when it starts going bad, but to be fair to Microsoft I'm pretty sure that's more the interaction of complex software with a massive amount of commodity hardware in configurations that are way to varied for all combinations to ever be tested. HP being especially bad and Lenovo especially good in this regard.
I do find myself occasionally wishing that the Surface range had the same hardware as my M1 Air as I'm still more productive on Windows, partly due to the UI/UX (when it is consistent which isn't everywhere) being better, and partly because it's just a whole lot easier to keep my hands on the keyboard and away from the mouse/trackpad than it is on Mac OS.
Whilst I do Python, Node, Go, Ruby, etc., I've been a .Net developer since 2001 and in many ways (going back to DOS in the 80s) Microsoft are responsible for me having a good career. I'm not blind to their rapaciousness, but I am grateful too.
"Python v3.9.1 becomes the first version of the language to support macOS 11 Big Sur. The developers note that the release is made possible thanks to Xcode 11..."
[1] https://www.techradar.com/news/you-can-now-run-python-on-app...
What would you like your computer to be faster at? If you just want apps to open faster in general, go for it. If you want faster performance for Xcode or video editing, I'd say wait until the end of the year.
https://www.apple.com/macbook-air/specs/
https://www.apple.com/mac/compare/?modelList=MacBookPro-13_1...
Snow Leopard dropped the price of the update from $129 to $29. This might have been part of its positive reception.
Apple’s design philosophy isn’t particularly different from that time. Snow Leopard was something of a feature freeze to make up for deficiencies with Leopard.
Apple was never some pariah of hardware and software flexibility. If you had a PowerPC Mac or older you dealt with all kinds of incompatibilities from different peripherals, floppy disk formats, and physical connectors.
The interoperability of Macs is the best it’s ever been, especially since Apple no longer deals with a <5% market share like the 1990s.
I don't think enough people appreciate just how beautifully hackable OS X used to be. The Library folder was unhidden by default, basically inviting users to go explore their systems. Applescript and Automator were treated as first-class citizens, offering an easy way to extend existing apps. Custom themes—while never officially supported—were easy to make via a bit of file replacement, and there seems to have been quite a lot of them, if DeviantArt is any indication. And there was nothing to stop you from writing and loading your own drivers to add support for new hardware, as plenty of enthusiasts did.
Today, stock apps like Music store data in obtuse and non-human-readable locations, Apple Events are virtually useless due to mandatory permission dialogs, kernel extensions are deprecated, and the root filesystem is mounted via a complex snapshot mechanism in which a reboot is required for changes to take effect.
Perhaps this was always Apple's goal. Perhaps it was necessary for security. Perhaps, for a majority of users, it ultimately provides a better experience. But if you are someone who likes to tweak and customize your machine, OS X used to be a great home, in a way that it isn't today.
I won’t address your whole comment but I feel compelled to point out that Music is iTunes. The Music app was just a name change, and, confusingly, Apple Music (the service) is an entirely different entity than Music (the app). Your own local music files (either your own MP3s or iTunes Store purchases) are still accessed in the same iTunes Media Library folder as they always were.
Music was just a rebrand, and it spun out podcast, audiobook, and iPhone/iPod sync. But the rest of it is 100% iTunes.
Could that be an indication that users don't really want new features in their OS?