Reasons for Buying an Apple Silicon Machine NOW(!) as a Xojo Dev

Obviously, there are many reasons for buying the latest (greatest does not yet apply) but I want to focus on the need for a Xojo Dev only.

Please don’t get me wrong: it’s the good ethical business practice to test our apps on all possible targets, but should ARM and Intel Apps not work absolutely identical on Silicone (in theory)? I mean will it not be sufficient (in this transition period of a few months until we will see real PRO Silicone devices) to test simply on an Intel Mac running Big Sur? What are your thoughts and arguments?

My view:

NO need to rush for Apple Silicon:

  • Web Development
  • RPI Development
  • Windows Development
  • iOS Development
  • console Development
  • macOS Intel Development (will run as an emulation on Silicon)

Apple Silicon recommended:
macOS Universal Binary (you might expect that the ARM will run identically, but it is better to test, but how big will the risk really be, I don’t know …)

Apple Silicon needed:
macOS pure Silicon ARM apps AND special function developed to be very “close” to the hardware. As you can only test such functions (via declares and/or with future plugins) it would be negligent not testing this on a Silicon machine.

I’m sure I have overlooked many good and valid arguments. What are your thoughts? I’m currently concentrated on iOS, Web, and Windows development, so I don’t feel in a rush. Users on macOS would currently be at max corporate users, who are not moving within the next months. But all that can of course change from one day to the other. But I’m usually cautious about whatever is “version/generation one” and I don’t think the current offerings are very appealing yet.

If you have a lot of macOS customers, you may just want to buy a MacBook Air and try your apps there to see if there are issues.

There will eventually be bugs only crashing on the Apple Silicon apps or only on the Intel apps and without testing hardware, you can’t verify them.

2 Likes

Actually I believe that anyone who has an iOS app that is successful enough to sustain one’s living, should test the app on macOS Big Sur with Apple Silicon.
The new Macbook’s and Mac mini can now run iOS apps, if the developer decides to make them available on Mac.

I expect three of my apps to get a few downloads per day on Mac so I definitely need to test them as soon as possible.

3 Likes

Granted. But let’s say one has only corporate macOS users using Intel. Now a few users might slowly move to Silicon. Do you expect they will see “specific” errors on Silicon only if we will deploy Intel App only and no universal binary yet? Of course, there will always be a risk, the question is how big this risk will be? ;-). Of course, you can’t seriously sell an app to the public, if not tested on Silicon. It is just obvious that we will see much better devices in a few months than right now. And on macOS you will probably need a Pro to test the Touch Bar behavior ;-).

Good point. Especially as the handling will be different with a mouse / keyboard. @Jeremie_L out of interest: will it be possible to test this in xCode on intel? Probably not …

No it won’t be possible to test this on Intel, and I don’t even know yet if xcode will add a window simulator to debug this.

Anyway, looking forward to receiving my new Macbook air :smiley:

1 Like

Sure in your and Christian’s case there is no other choice than being part of the game from day one and congratulations to the air: No fans is the killer argument for me. Even if there are better devices out one day, having no fans is always nice on a small device.

1 Like

My take is that Apple say intel compiled apps are expected to run under Rosetta2
So in theory, Mac apps built now should work as expected.

I dont really want to build bloaty UB apps.
So when Xojo can build for ARM, will we be able to build ARM-only?
If so, thats when we really need a device.
I may get one anyway, the Air looks tasty

Yes but… why? You’d rather build twice and hope your customers know which they have?

Yep, that’s my approach as well. The current devices aren’t bad at all (in contrary) for “normal” users or when someone has an urgent need for testing and running the business. But I’m missing all the new stuff which will now be possible with silicon without “attaching” apple chips to Intel chips (Face ID etc.). So it is very likely to see a lot more in a not so far future. On top of that my experience wasn’t the best in the past 2-3 years being an early adopter. Lessons learned ;-).

Yes, the reason why the download of Big Sur is > 10 GB :frowning:

I predict that Apple will be ahead of their roadmap for the complete move to ARM, as they were for the move from PowerPC to Intel.

But therefore they will need more “gadgets” for the mass, and it is one more argument for me to wait a bit. Once they will be adding face ID, Touch screens, or whatever else to Apple Silicon only, “ordinary” people will move. I don’t see this unique selling point for normal users yet. Only the “fan-free” Mac Book Air is appealing to someone not interesting in processors, chips etc. They just must come up with a lot more to get rid of the craziness of maintaining 2 major platforms - Apple hates it.

I am sure that one of the many Mac Mini in the Cloud providers will add Apple Silicone to their offerings pretty soon. This will be enough for me if I might run in the need to develop public apps for macOS. If one of my corporate customers wants to move in a rush, they need to provide me a device for testing.

How is that going to work with a Mini? If at some point I replace my Mini with a new one, I’m going to keep the mouse, keyboard, screen, that I already have. A touch screen is no replacement for a mouse, IMO - at least not so much that it becomes the obvious way to go. And my setup has no camera, which is another big plus for me.

Not at all, you’re right for this scenario.

But it is clear to me that for the mass market, Apple is targeting these “new” features (not sure about the touch screens (fruitless discussion for year now - I’m personally not missing them at all, but if you can run iOS Apps on macOS it makes sense to have such an interaction on notebooks). And the millenniums will soon approach things differently: if you can’t swipe it, it will be useless :-).

Beside the performance / watt challenge the biggest caveat for Apple was Intel not (or too slowly) delivering on new features. That’s why they had to add T2 chips etc. as a separate own chip and all the caveats related to this.

I have still a mini as well, and I’m using it only for testing. But a new device would most likely be a mobile device.

But I bet that we will see many new devices in 2021. COVID will probably screw up their schedule a bit. But with Apple’s aggressive 2 years plan, they have to deliver soon. And they will most likely target the mass market first (there is no Intel version for the Macbook Air any longer) as still will force the developers to follow anyway.

That’s the reason I won’t jump on the train in a rush. And more importantly, I’m fed up with the teething troubles of Apple lately. My new 16" MacBook Pro, bought in Dec’19, was useless for serious sound and movie editing for a few months.

I’m toying with the idea of getting the lowest end mac mini ($699 USD) instead of the Air just to use as a test machine, knowing that it will likely be a short lived machine. But it is a lower cost way to get a machine for testing for the short term.

Then in the future when I get a 16" MBP SoC, can retire it to a media server (with external disk) or Time Machine host or something. To be useful for more than testing, I’d want to bump up the Air specs and then I’m talking nearly twice the price of a low end mini. And still want a 16" MBP later.

I do wonder if iOS apps run on SoC handle multi-touch on trackpad/magic mouse better then the lousy quasi-multitouch on Intel simulators. Anybody with a DTK know?

1 Like

you can read my mind. Especially as the low priced MacBook airs are too pricey for what it is, and fully equipped the distance to a future MBP on SoC isn’t too far (or let’s say it doesn’t really matter then :slight_smile: ).

I do want to see some benchmarks on it, but maybe testing on the lowest common denominator is not all bad either. And I’d much rather do with a mini feeding another monitor input on my desk. My BT keyboard already switches effortlessly between up to 4 hosts. Wish the magic trackpad could.

But so far thinking that $699 USD for a temporary machine for testing may be my best bet, since I did not jump on the $500 DTK bandwagon.

Edit: Fixed typo of “with” to “wish” in first paragraph.

1 Like

Everyone debating this really needs to take the $699 plunge and grab that Mac Mini. It’s not going to be a short-lived option as it will already out perform most MBP and iMac systems. It’s a full bore Mac system. It’s not just an Apple TV with mouse and keyboard support.

One constant complaint is only 2 TB ports - remember these are TB 4 ports, so you get all of the performance of TB 3 with some great multi-channel support that was lacking in TB 3. Add a TB 3 dock like the OWC Thunderbolt 3 Dock:

Expand the connectivity dramatically by using one of the TB 4 ports on the Mac Mini.

Disclosure: - I now work for OWC. This is not my department, and I get nothing from it, but it’s still a great solution.

1 Like

I have a couple of OWC docks – love them.

My concern on the $699 unit is if 8GB is enough. On my 2018 mac mini (Intel…), I have 32GB memory and just starting Big Sur, XCode, and an iPod Touch simulator puts me over 16GB memory usage. I don’t have a DTK, so can’t tell what memory usage is like in SoC and if comparable or not.

What is memory usage like in your DTK?

The memory handling is quite different between the ARM and x86 architectures. As I mentioned in another thread, I’ve pushed my 16GB DTK far beyond the capabilities of my 16GB i7 MBP and the DTK still had breathing room. And, some of the apps on the DTK were still Intel running under Rosetta 2. Multitasking and task switching is far more fluid on the DTK, and the M1 is supposed to leave the DTK’s A12 in the dust comparably.

1 Like

Yeah, I figure a 16GB M1 mini will go a long way. But that extra $200 USB to go from 8GB to 16GB may be money well spent. (It also delays shipping at least in my area for 3 weeks or so already)

2 Likes

All good arguments, For sure these machines are great. My original question was more about, if we really do need it right now or if we can wait until the next generation. The latter is (currently) the case for me, as I’m currently only doing web dev. But it might change tomorrow.

My issue is that over the course of time I changed my habits considerably. A few years ago I had a whole farm of Apple devices - well I still have it but I’m not using it … For 2-3 years now I’m realizing that I mainly only work with my MBP. I still have an iPad Pro a Mac mini a iMac, but I don’t use them very often (mainly only to update them)… or not often enough for the price. So I’m all in for a new Silicon based MacBook Pro … but the 13" is not tasty enough yet (screen size) and yes, I’m still depending on a Windows in a VM at least once a week … So I will wait until a customer/project forces me to switch before the next generation.

But we can probably discuss this for an eternity, it is what it is. We are all different, we all have different requirements and Apple will always have the new “latest and greatest” :slight_smile: