Very Fast Transition

[pshufd]

I read a news article last night (sorry, don't have the link and I couldn't find it this morning), implying that the Apple Silicon transition will be very fast, similar to PowerPC to Intel. The article explained the reasoning for the fast transition - if you take your time, a lot of your Apps developers take their time. Articles, at this point, though are speculation.

My current setup is that I have a Mac Pro class Windows desktop that I just built and I really like it. I VNC into a 2015 MacBook Pro to run my macOS programs and this works out quite well. I can just take the MacBook Pro with me if I need to use it somewhere else. I'm looking forward to Apple Silicon but it's possible that it would be a secondary platform for me depending on price, performance, thermals and compatibility. I can live without x86 compatibility with my current setup because I have other x86 systems.

So, if Apple screws it up, it may push people off their platform, either fully or partially. If they come up with hardware that is just an order of magnitude better, then that's a different story. But it would have to be that good to go through the compatibility issues that I expect to see - that is in doing a complete platform instead of a hybrid platform.

 

[Yebubbleman]

I read a news article last night (sorry, don't have the link and I couldn't find it this morning), implying that the Apple Silicon transition will be very fast, similar to PowerPC to Intel. The article explained the reasoning for the fast transition - if you take your time, a lot of your Apps developers take their time. Articles, at this point, though are speculation.

My current setup is that I have a Mac Pro class Windows desktop that I just built and I really like it. I VNC into a 2015 MacBook Pro to run my macOS programs and this works out quite well. I can just take the MacBook Pro with me if I need to use it somewhere else. I'm looking forward to Apple Silicon but it's possible that it would be a secondary platform for me depending on price, performance, thermals and compatibility. I can live without x86 compatibility with my current setup because I have other x86 systems.

So, if Apple screws it up, it may push people off their platform, either fully or partially. If they come up with hardware that is just an order of magnitude better, then that's a different story. But it would have to be that good to go through the compatibility issues that I expect to see - that is in doing a complete platform instead of a hybrid platform.

With PowerPC to Intel, the roadmap was such that Apple was able to use the same Core Duo processors all across the line (as, back then, Apple was merely using the same mobile CPU for everything that wasn't the Mac Pro) and still have it outperform every G4 and most G5's. Then it was just a matter of switching the Power Mac G5 to a dual-core Xeon, which, conveniently came out that Summer, thereby kicking off the move to Core 2 Duo across the other Macs.

Today's Intel Mac range is far from being that simple. You've got 8th Generation Intel on three Macs (21.5" iMac, Mac mini, and 2-port 13" MacBook Pro), you've got 9th Generation Intel on the 16" MacBook Pro still, you've got 10th Generation Intel on the MacBook Air, 4-port 13" MacBook Pro, and 27" iMac, and then you've got different Xeon generations across both the iMac Pro and Mac Pro. The processor types even being used across those ranges are diverse (of those three 8th Generation Intel based Macs, you effectively have no fewer than two different KINDS of 8th Gen Intel CPUs; no fewer than three different KINDS of 10th Gen Intel CPUs; and the 9th Gen CPUs in the 16" MacBook Pro are yet a different KIND of Intel CPU relative to all of the other types being used in all of the other Macs). It's far from the simple "you have one kind of G4 in both the iBook G4, PowerBook G4, and eMac made by Motorola and then two kinds of G5s in the iMac and PowerMac/Xserve respectively" way things were done prior to the launch of the first Intel Macs.

Given that, the range of performance in each of these Macs ranges wildly across the lineup. The 8th Generation Intel based Macs, being more on the low-end can transition rather quickly. I'd imagine the MacBook Air (being on the really slow end of 10th Generation Intel) could also transition rather quickly. For everything else, Apple needs to produce chips for those Macs that outperform their Intel equivalent. I think some (4-port 13" MacBook Pro) will have an Apple Silicon counterpart quicker than others (16" MacBook Pro, 27" iMac, iMac Pro - if applicable, and Mac Pro).

Were it not for this wide range, I'd agree and suggest that it will all be over quickly. However, I think Apple will likely not transition the higher-end Macs over until A15-based SoCs start arriving. I do believe that the first wave (Mac mini, 13" MacBook Pro, MacBook Air, 21.5" iMac) will happen rather soon though.

Though, I also don't think that it won't be without some twists and turns (some Macs likely won't make the jump - 2-port 13" MacBook Pro and iMac Pro being most likely to fall in this category, while others may end up changing dramatically - 4-port 13" MacBook Pro may turn into a 14" MacBook Pro or just outright merge with the MacBook Air to become the one-size-fits-all sub-15" Apple notebook; we may even see some products get renamed as was done in the PowerPC to Intel transition).

 

[pshufd]

Though, I also don't think that it won't be without some twists and turns (some Macs likely won't make the jump - 2-port 13" MacBook Pro and iMac Pro being most likely to fall in this category, while others may end up changing dramatically - 4-port 13" MacBook Pro may turn into a 14" MacBook Pro or just outright merge with the MacBook Air to become the one-size-fits-all sub-15" Apple notebook; we may even see some products get renamed as was done in the PowerPC to Intel transition).

The article that I read had a main assertion that you need to transition quickly to get software developers to transition quickly. You have to set hard deadlines for stuff to go away and then it goes away. If you keep making hardware, then software developers, having lots of other things to do, will just tell customers to use the x86 hardware for now and they'll get to AS in their own good time.

 

[TylerL]

Today's Intel Mac range is far from being that simple. You've got 8th Generation Intel on three Macs (21.5" iMac, Mac mini, and 2-port 13" MacBook Pro), you've got 9th Generation Intel on the 16" MacBook Pro still, you've got 10th Generation Intel on the MacBook Air, 4-port 13" MacBook Pro, and 27" iMac, and then you've got different Xeon generations across both the iMac Pro and Mac Pro. The processor types even being used across those ranges are diverse (of those three 8th Generation Intel based Macs, you effectively have no fewer than two different KINDS of 8th Gen Intel CPUs; no fewer than three different KINDS of 10th Gen Intel CPUs; and the 9th Gen CPUs in the 16" MacBook Pro are yet a different KIND of Intel CPU relative to all of the other types being used in all of the other Macs). It's far from the simple "you have one kind of G4 in both the iBook G4, PowerBook G4, and eMac made by Motorola and then two kinds of G5s in the iMac and PowerMac/Xserve respectively" way things were done prior to the launch of the first Intel Macs.

This is more a side-effect of Intel thrashing about, trying to bring its new 10nm node into existence.
Surely Apple doesn't need to imitate Intel's strategy.
In fact, maybe Apple can simply differentiate (most) Mac product lines by TDP alone, reducing the number of unique CPU designs to tape out.

 

[Falhófnir]

This is more a side-effect of Intel thrashing about, trying to bring its new 10nm node into existence.
Surely Apple doesn't need to imitate Intel's strategy.
In fact, maybe Apple can simply differentiate (most) Mac product lines by TDP alone, reducing the number of unique CPU designs to tape out.

This is similar to what I expect, Apple will bunch computers into categories reflecting what a user will want out of it, and create a single chip for that market segment - as a pure example:

low power computing - MacBook Air, Mac mini = possibly the vanilla iPad A14X or if not an 'A14M' basic chip. About 7W, probably 4+4 cores.

competent allrounder - MacBook Pro 13", 24" iMac. About 15W (maybe TDP up on the desktop?) 6+4 cores.

performance centric - MacBook Pro 14", MacBook Pro 16", 27" iMac. About 30W (again probably with a bit of wiggle room on TDP/ clock speeds), 8+4 cores.

Professional - Mac Pro, iMac pro. Higher TDPs and core counts.

Like you say this might look a bit strange when you're used to the reality of Intel's product lineup (a 15W CPU in a currently 60W desktop chassis?) but if you can get 21.5" iMac performance out of a 15W TDP, and that is suitable for your target audience, why go to the extra expense of making a 60W overkill part?

I do also expect a certain amount of binning and clock speed based up-selling - so you might have, say, a 3GHz and 3.3GHz variant of a chip, and maybe a couple of core count configurations for the more powerful classes as well.

 

[ChrisA]

I'm not so sure transition speed is so important to developers. Let's say that today that all Macs offered for sale by Apple were Apple Silicon. Then tomorrow or even 6 months from today most Macs would still be Intel-based. If I wanted to sell software I'd make Intel software and I might wait a year to release my AS software.

As a software developer, I'd look at the installed base

The counter-argument is (I'm taking both sides as I said "I'm not so sure") is that I've read that people mostly buy the software right after buying a computer and then their rate of software buying goes down. SO a developer's customer base is biased to new computer owners. Still, it will take a LONG time until most Macs are Apple Silicon even if they finished the transition today.

 

[Jorbanead]

I'm not so sure transition speed is so important to developers. Let's say that today that all Macs offered for sale by Apple were Apple Silicon. Then tomorrow or even 6 months from today most Macs would still be Intel-based. If I wanted to sell software I'd make Intel software and I might wait a year to release my AS software.

As a software developer, I'd look at the installed base

The counter-argument is (I'm taking both sides as I said "I'm not so sure") is that I've read that people mostly buy the software right after buying a computer and then their rate of software buying goes down. SO a developer's customer base is biased to new computer owners. Still, it will take a LONG time until most Macs are Apple Silicon even if they finished the transition today.

This was true with the intel transition as well. Apple still made their transition within a year. Yes it will still take time for devs, but if Apple can transition their chips within a year, then we’re looking at about a 2-3 year transition period for developers (as the customer base slowly transitions). If Apple takes their time and uses the full 2 year window, then it’s more like a 3-5 year transition for developers (as there will be more intel overlap).

Honestly one of the main reasons for this is Apple wants to move towards an ASi-Only Mac OS but as long as they sell intel macs they’ll have to at least support OS upgrades for 2-3 years on intel (security updates for 5-7 years). If they can transition everything over in 2021, then by 2024 they can focus exclusively on their custom chips in Mac OS and not waist their time on dual support.

It’s still in Apples best interest to transition as quickly as possible. There’s also the issue of consumers (like myself) who will be in the market for a new machine soon, and I do NOT want to buy intel now because I worry that it’ll only get 2-3 years of support (not security updates, but OS updates). I’m just waiting for Apple to announce a Mac that fits my needs and for developers like yourself to transition over as quickly as possible.

 

[ADGrant]

The article that I read had a main assertion that you need to transition quickly to get software developers to transition quickly. You have to set hard deadlines for stuff to go away and then it goes away. If you keep making hardware, then software developers, having lots of other things to do, will just tell customers to use the x86 hardware for now and they'll get to AS in their own good time.

I don't think the article was written by a developer. There are 100 million Intel Macs in active use, Apple and third party developers will continue to support this user base for years. Also Apple will continue to sell Intel Macs for another two years or so, those Macs will not switch to vintage status for another 7 years.

2005 was a very different time, Macs had been shipping with OS-X pre-installed for only about three years and had been running on Intel CPUs before it was even released. I suspect that Mac OS-X on Power PC was always expected to be a transitional phase.

 

[ian87w]

I read a news article last night (sorry, don't have the link and I couldn't find it this morning), implying that the Apple Silicon transition will be very fast, similar to PowerPC to Intel. The article explained the reasoning for the fast transition - if you take your time, a lot of your Apps developers take their time. Articles, at this point, though are speculation.

My current setup is that I have a Mac Pro class Windows desktop that I just built and I really like it. I VNC into a 2015 MacBook Pro to run my macOS programs and this works out quite well. I can just take the MacBook Pro with me if I need to use it somewhere else. I'm looking forward to Apple Silicon but it's possible that it would be a secondary platform for me depending on price, performance, thermals and compatibility. I can live without x86 compatibility with my current setup because I have other x86 systems.

So, if Apple screws it up, it may push people off their platform, either fully or partially. If they come up with hardware that is just an order of magnitude better, then that's a different story. But it would have to be that good to go through the compatibility issues that I expect to see - that is in doing a complete platform instead of a hybrid platform.

I agree that Apple will do it quicker than their "promised" 2 years just because so they can yet again, claim to have the fastest transition in the industry. The 2 year promise imo was just marketing, Apple already planned a faster transition.

I think the hardware is ready, at least at base level. I'm more worried about whether big sur itself is really ready. The OS is key here, and I felt that even on a tightly controlled hardware like the iPhone, Apple is still having some bugs that annoy people.

 

[MikhailT]

Internally, Apple is way ahead of the public by many years. Apple planned out the transition several years ago, this isn't something they came up with a few years ago and you can tell by looking at what they've been doing with the software updates on both iOS/macOS over the past five years; stuff like AFPS, Swift, Catalyst, dropping 32-bit support, etc.

The problem isn't the hardware, Apple can easily swap everything out within a year. The problem is the software, they're clearly not capable of getting the software done in time. iOS 13 / Catalina was a massive flop in terms of software quality and they only managed to get their software QA down to acceptable level now (which is still not that great).

Even to this day, Apple hasn't ported all of their macOS apps to Catalyst (Books, etc) to showcase how well macOS/iOS can play together. Not to mention the first-gen Catalyst apps like News were horrible in Mojave (still is to this day in Catalina), which is why we haven't seen many iPadOS apps getting ported to macOS lately.

The success of the transition will depend on Apple's software quality and sustaining it for the next few years. Most experienced macOS users are probably going to accept the fact the first year will be problematic, the question is how fast Apple can address them all by the second year.

The pace of the hardware transition is being used to roll out software stuff in a methodical fashion. By starting with low-powered Macs like MBA, most users of that hardware aren't likely to run heavy apps like Photoshop, etc because of the hardware restrictions; these users are using more lightweight apps which won't give Apple any problems with Rosetta and give software developers enough time to port their heavier / complex apps.

 

[thenewperson]

iOS 12

iOS 13 you mean? iOS 12 was beloved, and also came with Mojave.

Even to this day, Apple hasn't ported all of their macOS apps to Catalyst (Books, etc) to showcase how well macOS/iOS can play together

I can't imagine they'd want to do this when they don't need to. It's likely we'll get Catalyst versions of apps needing redesigns based on their iOS counterparts (so I do expect a Catalyst iBooks soon-ish), but for everything else it's likely going to be AppKit until they can do a SwiftUI version.

 

[pshufd]

iOS 13 you mean? iOS 12 was beloved, and also came with Mojave.

I'm on Mojave and will probably stay on it for at least another year.

 

[theluggage]

The article that I read had a main assertion that you need to transition quickly to get software developers to transition quickly.

They've said "about 2 years" which, I think, is the same as they predicted for the Intel transition. Since they've semi-committed to that it probably includes a bit of slack (always double your time estimate!) They ended up completing the Intel switch in 1 year, which probably means that they got lucky (although they didn't have a pesky viral pandemic to gum things up then). It was also under 2 years from Mac OS X public beta to 10.1 becoming the default OS on new Macs - and in many ways transitioning to a completely new OS where nothing "just re compiled" was a tougher call on developers.

So the report you read wasn't exactly going out on a limb...

 

[dmccloud]

There is one thing I've been contemplating with regards to both the timeframe for the transition and which models will get updated to AS first. During the WWDC keynote, Apple made sure to highlight that both Microsoft Office and Adobe software was already running on Apple Silicon. While the Office side of the equation makes sense just out of the sheer volume of Office users on the Mac, the Adobe component of that would really be targeted towards those Mac users who rely on the Adobe suite on a daily basis. That's not going to be your average MBA/iMac customer, but the Mac Pro/MacBook Pro/iMac Pro user base. So I'm wondering if we may see something along the "Pro" spectrum coming out in the first batch of Apple Silicon Macs, perhaps just as a means for Apple to showcase the potential of the platform.

 

[DHagan4755]

It'll actually be good for them to move to their own silicon as quickly as they can.

 

[ADGrant]

They've said "about 2 years" which, I think, is the same as they predicted for the Intel transition. Since they've semi-committed to that it probably includes a bit of slack (always double your time estimate!) They ended up completing the Intel switch in 1 year, which probably means that they got lucky (although they didn't have a pesky viral pandemic to gum things up then). It was also under 2 years from Mac OS X public beta to 10.1 becoming the default OS on new Macs - and in many ways transitioning to a completely new OS where nothing "just re compiled" was a tougher call on developers.

So the report you read wasn't exactly going out on a limb...

The viral pandemic does not seem to be slowing them down. What may slow them down (aside from the size of the Intel Mac ecosystem) is the breadth of Intel's CPU range installed in Macs. Everything from whatever it is that Apple puts in the MacBook Air to a 28 core Xeon.

 

[vigilant]

With PowerPC to Intel, the roadmap was such that Apple was able to use the same Core Duo processors all across the line (as, back then, Apple was merely using the same mobile CPU for everything that wasn't the Mac Pro) and still have it outperform every G4 and most G5's. Then it was just a matter of switching the Power Mac G5 to a dual-core Xeon, which, conveniently came out that Summer, thereby kicking off the move to Core 2 Duo across the other Macs.

Today's Intel Mac range is far from being that simple. You've got 8th Generation Intel on three Macs (21.5" iMac, Mac mini, and 2-port 13" MacBook Pro), you've got 9th Generation Intel on the 16" MacBook Pro still, you've got 10th Generation Intel on the MacBook Air, 4-port 13" MacBook Pro, and 27" iMac, and then you've got different Xeon generations across both the iMac Pro and Mac Pro. The processor types even being used across those ranges are diverse (of those three 8th Generation Intel based Macs, you effectively have no fewer than two different KINDS of 8th Gen Intel CPUs; no fewer than three different KINDS of 10th Gen Intel CPUs; and the 9th Gen CPUs in the 16" MacBook Pro are yet a different KIND of Intel CPU relative to all of the other types being used in all of the other Macs). It's far from the simple "you have one kind of G4 in both the iBook G4, PowerBook G4, and eMac made by Motorola and then two kinds of G5s in the iMac and PowerMac/Xserve respectively" way things were done prior to the launch of the first Intel Macs.

Given that, the range of performance in each of these Macs ranges wildly across the lineup. The 8th Generation Intel based Macs, being more on the low-end can transition rather quickly. I'd imagine the MacBook Air (being on the really slow end of 10th Generation Intel) could also transition rather quickly. For everything else, Apple needs to produce chips for those Macs that outperform their Intel equivalent. I think some (4-port 13" MacBook Pro) will have an Apple Silicon counterpart quicker than others (16" MacBook Pro, 27" iMac, iMac Pro - if applicable, and Mac Pro).

Were it not for this wide range, I'd agree and suggest that it will all be over quickly. However, I think Apple will likely not transition the higher-end Macs over until A15-based SoCs start arriving. I do believe that the first wave (Mac mini, 13" MacBook Pro, MacBook Air, 21.5" iMac) will happen rather soon though.

Though, I also don't think that it won't be without some twists and turns (some Macs likely won't make the jump - 2-port 13" MacBook Pro and iMac Pro being most likely to fall in this category, while others may end up changing dramatically - 4-port 13" MacBook Pro may turn into a 14" MacBook Pro or just outright merge with the MacBook Air to become the one-size-fits-all sub-15" Apple notebook; we may even see some products get renamed as was done in the PowerPC to Intel transition).

I think there's a couple of things that could simplify this a great deal. First, it seems clear to me by watching the frameworks discussions at WWDC for the past few years, as well as Platform State of The Union, that Apple has bee laser focused on pushing people to Apple specific purposes. Additionally there was a push off of 32 bit.

Theres a lot of variables, and I don't know if we have a full picture yet of what Apple is doing. I'd LOVE to see a 6 by 6 design on a MacBook Pro between energy cores and performance cores. I'm sitting here on a new MacBook Pro 16, and it's embarrassing that running a multi video Teams call can make this drag *** and kick off the fans as loud as possible.

Apples frameworks target GPU, FPGAs, CPU, and "Neural Engine". Lets build something competitive. Active cooling is totally fine and cool on a MacBook Pro. I expect it for "hard" tasks, not conference calls.

 

[ADGrant]

I think there's a couple of things that could simplify this a great deal. First, it seems clear to me by watching the frameworks discussions at WWDC for the past few years, as well as Platform State of The Union, that Apple has bee laser focused on pushing people to Apple specific purposes. Additionally there was a push off of 32 bit.

Theres a lot of variables, and I don't know if we have a full picture yet of what Apple is doing. I'd LOVE to see a 6 by 6 design on a MacBook Pro between energy cores and performance cores. I'm sitting here on a new MacBook Pro 16, and it's embarrassing that running a multi video Teams call can make this drag *** and kick off the fans as loud as possible.

Apples frameworks target GPU, FPGAs, CPU, and "Neural Engine". Lets build something competitive. Active cooling is totally fine and cool on a MacBook Pro. I expect it for "hard" tasks, not conference calls.

Well you are using MS Teams which is one of those Electron web apps I believe.

 

[vigilant]

Well you are using MS Teams which is one of those Electron web apps I believe.

Yeah that's fine. Webex, and Zoom do it too. Webex specifically caused my laptop to just completely shut down. I pulled it up on my iPad and had no issues. It was a 2 hour AHOD, but I couldn't get it back up and running without turning off video for all participants.

Zoom, it does go full on fan as well.

 

[JMacHack]

There is one thing I've been contemplating with regards to both the timeframe for the transition and which models will get updated to AS first. During the WWDC keynote, Apple made sure to highlight that both Microsoft Office and Adobe software was already running on Apple Silicon. While the Office side of the equation makes sense just out of the sheer volume of Office users on the Mac, the Adobe component of that would really be targeted towards those Mac users who rely on the Adobe suite on a daily basis. That's not going to be your average MBA/iMac customer, but the Mac Pro/MacBook Pro/iMac Pro user base. So I'm wondering if we may see something along the "Pro" spectrum coming out in the first batch of Apple Silicon Macs, perhaps just as a means for Apple to showcase the potential of the platform.

Ehhhh, Apple sells truckloads of 21" iMacs to Universities for Graphic Design, and MacBook Pros to Students. That's not necessarily highest of the high end, but fits well into the desktop publishing demographic (which is probably Apple's most devoted base). The low-end iMac and laptops I could see a quick rollout. It's the higher end stuff that I'm not sure they have ready.

iMac 27", iMac Pro (assuming they keep it), and Mac Pro are probably gonna be the last to make the jump. They need more than raw CPU performance (though that is very important), they need great GPU and I/O as well. Moreso with the Mac Pro.

Also, recent trends have shown that whenever Apple promises a release date "in a year" it's likely gonna be November/December. See: Mac Pro 7,1, The latest Mac Mini, iPhone 12, etc. I'd bet that the first ASi Mac won't ship until December, and the final Mac to make the transition (The Mac Pro) won't ship until December 2022.

I'm on Mojave and will probably stay on it for at least another year.

I use Catalina at work, and it's a buggy mess. Mojave is easily the best release of Mac OS. (fite me)

 

[Yebubbleman]

I use Catalina at work, and it's a buggy mess. Mojave is easily the best release of Mac OS. (fite me)

I'm not going to say that these are better, but I did find that Mountain Lion and El Capitan were about as good. Though none of those three compare to the grandpappy of stable releases, Snow Leopard.

 

[ADGrant]

Yeah that's fine. Webex, and Zoom do it too. Webex specifically caused my laptop to just completely shut down. I pulled it up on my iPad and had no issues. It was a 2 hour AHOD, but I couldn't get it back up and running without turning off video for all participants.

Zoom, it does go full on fan as well.

Apparently it is not fine if your MacBook is struggling. On the iPad, the app may be Native so not really a fair comparison.

 

[MikhailT]

iOS 13 you mean? iOS 12 was beloved, and also came with Mojave.



I can't imagine they'd want to do this when they don't need to. It's likely we'll get Catalyst versions of apps needing redesigns based on their iOS counterparts (so I do expect a Catalyst iBooks soon-ish), but for everything else it's likely going to be AppKit until they can do a SwiftUI version.

Yes, I meant 13. Updated my post.

As for Catalyst, they need to dogfood their stuff. Apple cannot expect developers to report bugs and find issues with their frameworks. The more the company test the transition to Catalyst, the more polishing they can add quickly.

In certain WWDC videos, they actually came out saying that they saw problems when they converted their apps and so on.

 

[thenewperson]

As for Catalyst, they need to dogfood their stuff. Apple cannot expect developers to report bugs and find issues with their frameworks. The more the company test the transition to Catalyst, the more polishing they can add quickly.

I agree with this, I was just disagreeing with the idea that they needed to port "all" their apps to Catalyst. They should port the ones with iOS counterparts which are much ahead at first, which seems like what they're doing now. No need to touch the many system apps with no iOS equivalent right now just to show that they believe in Catalyst. I feel like Messages, Maps, Podcasts, Find My (and maybe even future Books, iWork, and Photos ports) are okay at this time.

 

[JMacHack]

I'm not going to say that these are better, but I did find that Mountain Lion and El Capitan were about as good. Though none of those three compare to the grandpappy of stable releases, Snow Leopard.

Having used Snow Leopard at work, try leaving a .pdf file highlighted in Finder, then exporting a QXP file as that same .pdf. Finder crashes every time, corrupting the .pdf. I've lost count over how many times this happened to me, my blood pressure still hasn't recovered. Plus QXP 8.1 crashed all the goddamn time, it's easy to run things in Rosetta without noticing (I did!).

Plus there's little things like window snapping and resizing being much improved (dunno which release this happened), spotlight being better, etc. Searching in Finder seems to have regressed somehow though. Spaces is inferior to the later multi-desktop implementation imo.

El Cap is where I started, and it was good. Was it Mountain Lion or Lion that had the bug where it would eat up all the available RAM? I remember that bug hitting several school computers, making them slow because everything was constantly swapping. That was a headache.

Mojave has APFS, the dynamic desktop (which I think is cool!), and stacks, which surprisingly came in handy more than I thought.

I dunno, I'm a big Mojave fanboy. I don't think Snow Leopard aged that well, I tend to lump it in with mid-classic OSX (Panther, Leopard)