Phil Schiller, Craig Federighi and John Ternus on the State of Apple’s Pro Macs

"Do you use Final Cut Pro?" No, I use scikit-learn! No, I use vagrant & virtualbox to run 4 VMs No, I use Unity ...

Until this announcement, I assumed, based on the neglect of the mac pro and mac os, that Apple was basically going to just run out the mac business. I had started trying to figure out what my next computer was going to be after a decade of macs, and was pretty bummed by the options.

Hopefully they're serious about this.

If they're looking for other things they could do, may I suggest doubling down on fixing mac os quality issues. And hey, maybe even hire the homebrew folks.

I also don't quite get what are they going to be spending all this time on, what kind of "designing" is going into it? Picking components on NewEgg and figuring out the shape of their Apple-branded case? :)

Seriously, unlike mobile/laptops, Apple has no competitive advantage in traditional desktop computing. They don't make CPUs, chipsets, GPUs, etc. And there's no battery to fiddle with, or a case to make thinner or a keyboard to mutilate.

[1] https://www.tonymacx86.com/buyersguide/april/2017#CustoMac_P...

I've never quite figured out the continued popularity of the ESC key for vim using developers. For a touch typing vim user on most modern keyboard layouts generating an ESC by hitting the ESC key is more movement than generating an ESC by hitting CTRL-[. This is especially so if you have a keyboard that puts the CTRL key where God intended it to be (immediately to the left of the 'A' key). With CTRL in that location, hitting CTRL-[ is just a small pinky movement on each hand away from home position.

In earlier times, ESC made a lot of sense. On the ADM-3A, for instance, CTRL was in the proper location next to 'A', and ESC was immediately above CTRL, and '[' was at the far right side of the top row (near where BACKSPACE/DELETE is on most current keyboards). On that layout ESC was way more efficient than CTRL-[.

I am pretty sure it's still a very very small percent out of all the MacBook Pro buyers, and a lot of the vim users are quite happy with just remapping the each key and were doing that even before they got the touchbar. I think this whole buzz about missing esc key and fn row is just non relevant anymore and the vast majority of MacBook users myself included don't really care and enjoy a new great screen along with a better trackpad and battery life.

I never understood how this silence helps any company (except in cynical way - we won't tell them when they'll get new model and how good it will be so they will clear our stock of soon to be 'old' models).

You are VASTLY underestimating the planning/engineering/marketing/design/fabrication/logistics/etc. that a multi-billion dollar company undertakes to bring a new product to market, especially one whose brand is built (in large part) on design. Just rolling out something we've seen before isn't going to happen and even then it's not something that could happen quickly.

It's really just that simple.

Apple pundits are always talking about how Apple has prototypes of everything imaginable under the hood that they won't release if they don't think it's just right. And yet they were so blindsided on the Mac Pro that they didn't start developing a modular one until now?

Apple's earlier iphone & ipod cables are a more common example of this. They broke regularly because they were built to a design spec, not an engineering spec.

I think you're making the oversimplification they raise in the article about grouping all of the different pro users archetypes into the idea of "the pro user". (Some) developers might not care, but other pro types especially that make heavy use of peripherals certainly do.

Yes, it's easier to do all of this with a bigger footprint, but Apple has a hard time pleasing people with the Mac Pro regardless of which direction they go, and at this point I don't think they would win over many fans by pushing out a standard tower.

Lets say it takes 6 months to design and test a new system at peak design efficiency using current parts. If you want a cutting edge system at launch though you need to design and test it to take advantage of the CPUs, GPUs, chipsets, connector technologies etc that will be cutting edge then, not now. That’s a lot more complex and takes much longer. Also the launch schedule is not just driven by your own product development, it also has to be synchronised with the component manufacturer’s product development. That can easily push out the schedule even further because you’re not working with current known, well understood parts. You’re working with pre-production specs and early prototype samples.

Most manufacturers don’t get privileged acces to these, but you can bet your bottom dollar Apple does. Apple also have a history of collaborating with parts manufacturers to specify custom variants of parts or even designing entirely new parts, such as the custom timing controller chip in the original 5K iMac. Working at that level takes even more time.

ATP guess around 200,000 Mac Pros are sold a year - the first year of a new version after so long could easily be 1 million units. Getting that production and logistics line set up is a big undertaking even if using old cases etc. For the extra 6-9 months that a new design would take it's totally worth it for them as will hopefully give them a modern design that is good for 5-10 years that meets the needs of the 2017+ pro market, using a chassis that was designed for the 2008+ needs would be a step back for Apple and pros in general. Apple like to push the envelope and while they certainly do make mistakes (I count the trash can as a mistake but not the Touch Bar) there's no point in them going ultra conservative.

Having said that, it'd be fun to see a Cray-like thing that wraps the PCIe boards around a flex-PCB bus going around a central core (CPUs, memory, heat management) while keeping the mass storage between the empty wedges created that way.

The fashion statement as you put it is their biggest differentiator. You are being unfair or shortsighted from their perspective.

1 year is pretty amazing.

One can argue that current iphone/ipad is not future proof due to non-usb-c port, for example.

Why exactly? Sure there are some people who want legacy connectors, but majority don't really care, in the last 3-4 years the only time I used non type c USB connector is when I was charging my phone. When I got a new Mac I just got a type c cable for 10$ - problem solved.

Edit: it is amusing to see that some think that random guy on the internet is better informed than the executive who sells the product and has all the data.

I think the transition was complete with the release of the iPod and iTunes and the way they advertised it.

CFD guys are still using software from the 80s and 90s. Same for finite element guys. Video producers already know whether they prefer Adobe premiere or final cut etc.

Flashy new release titles matter much less than specs, I'd say. That, and upgradeability.

"But can't everyone just implement our Metal API? It's been so successful, it's the best, you'll love it" No. No we don't love it. Please just implement WebGL2. Thanks.

"But we neglected OpenGL for years, even though our pro users have been begging for updates, and we think Metal is the best, you'll love it".

No. We want OpenGL updates. We are pros, and you promised us for years OpenGL updates. This is why there are almost 0.0% of our apps released on Macos with Metal.

"But we released the new macbook pro with an AMD card. We haven't really been working on OpenCL, but use that. You'll love it".

No, we want CUDA. We are pros, and we told you we wanted CUDA. It's what we use, because you stopped caring about OpenCL.

"I know we said we were serious about pro models, and that there would be a new mac pro, but here is the same one with a minor update. You'll love it. It's the best."

No.

"But we're listening."

No, you're bloody not.

...the link is to a full transcript of the meeting. Sure, you weren't personally invited, but hardly a secretive off the record briefing.

> Steve always ran things incredibly tight lipped, no need for transparency that's part of the Apple experience

After the controversy about the iPhone 4's antennas he did a very similar briefing to this with a number of reporters, and Apple pre-announced the Intel transition many months before it actually happened. Jobs would happily tell you what was coming down the line when it suited him to do so.