ARM Mac Part Deux: Less Confusion | Hacker News Reader

ARM Mac Part Deux: Less Confusion

mondaynote.com

41 pointsbookofjoe45 comments

To me the fundamental difference between a PC and a tablet is the software ecosystem: can you install any program/OS or must you install only those approved by the platform owner; can programs access system resources liberally (including a shared storage namespace) or are they tightly sandboxed; how much of the underlying infrastructure is visible to and manipulable by the end user. Compared to this difference the question of whether the device has an integrated keyboard or not seems minor to me.

Moving from a PC like ecosystem to a tablet like one is a net loss in my opinion. People are willing to absorb this loss and enter into a walled garden because right now building iPad-like hardware is not easy. I hope it doesn't continue to be so in the future.

Can you go more into why the difficulty of creating the hardware leads to willing adoption by users that most likely will never see "how the sausage is made"? I would think it would be the opposite (commodity hardware and easy app creation would lead to wider spread adoption by the hobbyist crowd).

I think it will mostly boil down to what you grew up with. I'm 41, I came of age building my own PCs, getting excited about an OS (Win95), the tools of creation (the PC) differing from the tools of consumption (the smartphone/iPad). My son and daughter (3 and <1 yrs) will be exposed to PCs due to my job, but I bet a lot of their friends in daycare will see a tablet as "the way computing has been and always will be". (Honestly my kids may think that too, daddy just hasn't accepted how much of a dinosaur he is yet).

Just like in our generation there will be people in our kids' generation who will want to know how the sausage is made and it will be a tragedy if the only way to do so would be to grow up, join a big firm and sign a NDA.

Even for the others, an open platform means a greater variety of software for end users. Creating toll gates on the way to software development or suppressing software which reduces the platform owners's profit is not good for end users. Many people who will never touch a command line still benefit because a Finnish computer science student could decide to start writing a PC operating system of his own and that OS now powers their phone or allows others to provide online services on servers that runs that OS.

If building iPad clones with an open hardware specification like the PC's becomes possible for multiple vendors I'm sure we will see a software ecosystem that will easily beat the walled gardens. The problem is that currently the number of firms that can build the hardware is small enough that they find it in their self interest to use closed specifications.

Similar age group here.

As you remember the PCs were the outliers, everyone else was mostly offering the same hardware/software integration as tablets.

In fact Apple is the surviving company of those integration and most OEMs what to get back to it, because they hardly make any money nowadays just selling parts.

Had it not been for Compaq and a failed lawsuit, building our own PCs would have been an alternative universe, which I never bothered to do anyway (was more expensive than buying a full new PC on my region).

I'm not sure there's really a contradiction there: "the hobbyist crowd" is, and has always been, much smaller than the general population, so even with commodity hardware and easy app creation leading to widespread adoption within that subgroup, you can still have more people overall wanting, buying, and using something more like an iPad.

As I am trying to put my iPad Pro 12.9 to more productive use, the single most limiting thing is, that the files application does not have access to all apps document storage. E.g. I cannot move a movie file from iCloud to the TV app or vice versa. Nor can I easily move other files from app to app. Nor can I copy files from and to external media connected via USB. Images are special cased as apps can access the photo library and you can drag and drop them, but real file access via the files app would be more flexible. Fixing this should be possible without a conceptual change of the app concept and would enhance the productive use immensely.

As a developer of course there is the other big limitiation: the constraints of developing software on the iPad. For me, this is the main distinction between a device and a computer: the ability to program it. Due to the app store and its rules, iOS is a very secure and stable environment. This makes the iOS platform very attractive, and I think it was the right approach to start with a restricted environment and remove some restrictions over time. So I do think it makes sense to limit distribution of apps to the app store, but the restriction on what apps can do on the device should be reviewed, even if that blurs the lines to the app store. An environment like termux, which still could be completely sandboxed, would increase the usability of the iPad a lot. Or even, going completely crazy, a whole Linux VM in an app sandbox. On the other side, having a full blown IDE for creating and locally running iOS applications would be a huge step forward.

The hardware itself leaves little wishes - except perhaps for a small overwork of the smart keyboard to have that additional row of function keys (less urgent) and of course the escape key. And little improvements of the ergonomics of the cursor and the return key. While the non-programmers users probably can live with the current layout, I am really surprised that Apple is ignoring the needs of any programmers so thoroughly.

>On the other side, having a full blown IDE for creating and locally running iOS applications would be a huge step forward.

Are you aware of iOS Dev environments such as Pythonista? It’s even possible to take a Pythonista app, copy it into an XCode template and release it on the App Store, though of course that has to go via a Mac.

To me it seems like the traditional PC model has too many downsides in terms of cost to support applications from a developer perspective as well as inferior security in comparison to a more sandboxed model. For most users, a walled garden is better than an open software ecosystem. That being said, I think the best hope we have for an open ecosystem at this point is web applications. They don't have the same downsides as the traditional PC software model. The sandboxing you get with web apps may restrict some of the functionality you could get relative to traditional PC software, but IMO the security benefits are far more important. The risk here is that these all end up being tied to centralized cloud services owned by a handful of giant tech companies, and that's ultimately not as open as we'd like. But it's at least a significant improvement over the App Store model.

A web site is running on someone else's hardware or running on your hardware but dependent upon others resources to operate. In theory you could open the websites sources but since the goal of app stores seems to be finding ways to monetize peoples data or get them to pay you monthly the end result will probably be unfriendly to tinkers, unfree, and closed.

This generation will be poorer and dumber than the last.

You can still have a sandboxed model with having the root user accessible, companies don't build that because they don't want the user to have any power on their machine but there's no technical limitation to do it.

I don't know if it still is the case, but in the early OS X adoption days Apple forums were full of people asking how to just run as root.

So....

Of course you will always have people to try to run things as root but the option could just be hidden like the developer options on Android with a big warning that enabling root might destroy their phone.

Offtopic, it's fun to see french idioms creep in to Gassee's writing. "Gas refinery" is a direct and ill-fitting translation of the french idiom "usine a gaz", which refers to something over-engineered and inefficient. Like a factory that's only producing hot air.

Edit: and it's likely Gassee knew exactly what he was doing by using that term, since that article includes a paragraph about domain-optimizations that don't translate well to others, specifically calling out certain foreign words that have no direct translation. Whoosh to me...

The "cars" vs "trucks" analogy is interesting, considering that in US, SUV's (a category that itself includes a sort of hybrid car/truck) outsell cars with many saying the "sedan" is dying [0]

[0] https://www.bloomberg.com/news/features/2018-01-16/why-the-a...

I believe that jobs was referring to a 'real' truck like a semi truck.

More likely a pickup. Still, that's a small % of the overall market, which is mostly headed in the direction of hatchbacks, crossovers, and SUVs. The reason Ford is staying in the truck market but greatly de-emphasizing it's sedans is that their trucks are generally considered to be the best you can get, and with their sedans its the opposite. It would be a major investment for them to try and compete with everyone in that segment even if it weren't shrinking. Although for what it's worth, their newest crossover, the EcoSport, is basically just a taller Fiesta (arguably their best sedan/hatchback despite it being the cheapest).

Anyways, in many regards the growing segment of the market is more like sedans/hatchbacks on stilts than it is like trucks.

Also, pickups most probably wont ever be "designed" to be self driving cars. They may have that has an option in the distant future. But given they are (supposed) to be farm/woodland/workhorses vehicles, they will stay mostly as they are for a very, very long time and give Ford a century to adapt (or milk the cow, kill the cow, eat the cow and close shop). The US/Canada is full of thousand of miles of woodland tracks and farmland unpaved tracks self driving vehicle will never be able to map. So unless they can drive fully on instruments including snow and very close trees. People really do live at the end of these tracks (at least in the Appalachians) and owns pickup trucks just to get home (and the status symbol thing).

As for the pickups as "status symbol", it's irrelevant because the very symbol is the offroad "real work, real duty, real matcho mens" factor. Given Ford seems (as an outside observer) not to care about electric and self driving, focusing on bigger Lincolns and the F-series makes total sense as a business plan.

1) Let the other spend their billions developing their hipster tech

2) Use reliable incomes that will outlive any trend

3) Get mature technology on the cheap once it's commonplace.

4) Ignore "carmaker has a service" because it's not their core competence and there will be profitless races to the bottom by the "data warehouse" players (rather than profit for the carmakers)

5) Profit

Almost all trucks are used like cars in the same environments 99.9% of the time with the added ability to haul things.

The use case you describe is best served by having a manual override button and retaining a traditional steering wheel as opposed to having a different more manual product

True, but that's not the marketing they want. And the market segment that buys these things are very sensitive to this mythical image. "True" trucks is a selling point, not a liability for that crowd.

When they sell a pickup to someone who doesn't need one (>85% of the sales), they sell the myth and the social status. So, as my last paragraph pointed out, they will but "not self driving" as a "status symbol" and they will sell like hotcakes.

Plus, the advantages of self driving for cars you own, considering the current price premium, wont be worth it for many, many years. For cars you /rent/ fine, they will "work" 24/7, but spending 15k-25k on LiDAR and sensors and data and engineers and (carmaker) insurances isn't worth it yet for a car that spends 98% of it's useful life parked.

25k worth of lidar makes it worthless for most market segments I think. Lacking self driving features will ensure that the insurance costs 3 times as much ultimately. It will eventually be cheaper to have self driving capability than pay premiums.

> pay premiums.

When you buy the self driving car, the manufacturer has to put a lifetime insurance on the car. It's baked in the retail price unless they decide to rent you the right to use your own car (that wont be good press). So actually, you pay more for the insurance in the hidden fees. In that imaginary 25k there is:

* The hardware $$

* The engineering (past, present, lifetime codebase maintenance) $$

* Recalls because the LiDAR turn out of miss important things $$$

* Maintaining the mapping data forever $$$

* The forever/lifetime insurance if the car kill someone $$$$$$

* That thing called profit margins $ (or no $ at all)

* Buyback cost (because they really can't support it forever) $$

That is, of course, if they do the "right thing" and contractually agree never to choose to turn your car into a very large brick (agreed buyback if they decide to retire the tech). This assume a real self driving car, no steering or fallback.

Selling high tech full self driving cars seems like madness. Lane tracking and "cheap" camera tricks with "the driver has to be alert all the time, if he dies then go blame his corpse" are the only way to turn a profit on driving assistance techs.

Do you expect accident rates among manually-driven cars to substantially increase? If not, why do you expect insurance premiums to substantially increase?

I expect accident rates to fall due to safety features but ultimately expect almost everyone to stop driving their car and pay relative to now much less for insurance. As the safety of the majority rests largely on getting the remaining idiots off the road i expect them to pay punitive insurance rates not dissimilar to sin taxes discouraging behaviors that are disadvantageous to society.

Im short a small group of people requires a greater cost to profitably insure plus sin taxes for insisting on endangering your fellow citizens vs falling rates for everyone else.

I actually really liked Apple's iPad Pro keyboard.

It's a heck of a lot better than the MacBook Pro one which really has to be Apple's worst technological innovation in the last 20 years.

This is so crazy to me because I love the MacBook Pro keyboard (assuming you mean the one on with the Touch Bar). I know it's personal opinion and there are tons of people that don't like it but I seriously have no issues with it and feel like typing on it is just a joy. It's best keyboard (save for my mechanical Ducky) that I've typed on. It's definitely my favorite laptop keyboard, even more so than my 2011 MBP and 2013 MBP.

Agreed. Have learned to love the very short travel keyboard of my Touch Bar MacBook Pro. I note that my typing is more effortless now somehow.

It was a little odd to start but I've also really grown to like it, and now find typing on the old one strangely clunky. The smaller travel of the keys and the reduced key height actually helps me type a lot faster.

I think some people are never going to feel right typing with a "thin laptop" range of key travel. I can deal with it, but it still feels weird.

The new mac book pro keyboards are really good to type on, satisfying and fast. Problem is, some keys stop working properly after a while (lookup the classic b key - stuff gets jammed in it!)

But I would still call the new MBP Apple's worst keyboard because, the arrows are totally useless. If you type emails all day -> great. If you're a developer -> rubbish (yayaya I don't do VIM).

If you did use vim, and had the touchbar, it would still suck because the esc key isn't a real key anymore.

Apple added a built-in option to map the caps lock key to escape. This not only makes sense from a RSI and efficiency perspective but is also closer to the keyboard that vi was originally designed for (the escape key was where modern PC keyboards put the tab).

A lot of dedicated vimers will use ctrl-[ instead of escape to avoid the inanity of reaching up there regardless of their keyboard, which is also a good approach if you're in the camp that maps caps lock to ctrl.

http://vim.wikia.com/wiki/Avoid_the_escape_key

I'm not a heavy vim user, but this is THE reason I was afraid of the touch bar. It turned out not to be an issue. I honestly don't think even heavy users would have much of an issue.

My only problem with the touch bar now is that it has crapped out on me more than once requiring a restart.

If only you remapped Capslock to Esc like 80% of the vim developers out there...

Toggling the capslock LED by using capslock as esc would annoy the heck out of me.

The caps lock LED stops toggling after remapping it to escape on a mac.

I might try that, then.

Everyone knows real vimmers map CAPS LOCK to esc.

Maybe this whole thing was just a big troll by the Apple engineers to coax more people into doing it "the right way" ;)

It's sort of like the nerdier version of the Jobsian "you're holding it wrong".

Your fault for not using vim :)

Its replaceable as well if your macbook pro keyboard flakes out you're in a world of hurt.

No doubt, since it cannot be fixed at the store, but they don’t tell you that until your already at the appointment. Sending something in because of a broken key is just poor design especially for a company that touts being environmentally friendly.

My touch bar 15” MBP keyboard was replaced on site (took about 2 hours) at the Portland, OR Apple Store.

You seem to be getting downvoted here. I agree with you: the Smart Keyboard is fantastic to type on, imo. And it’s impervious to spills and crumbs.

My biggest beef with Apple laptops is how tiny the return key is on European keyboards.

My bet is they will revive the iBook brand.

Interesting to note the author’s role in the Newton: https://gizmodo.com/5452193/the-story-behind-apples-newton

Reply on news.ycombinator.com