Ask Paul: March 8, 2024 (Premium)

Happy Friday! This is my final Ask Paul from Mexico on this trip, and it’s chock-full of excellent reader questions across a wide variety of topics. Let’s dive in.

Self-service is the future of hardware

thewarragulman asks:

With more and more PC manufacturers such as Lenovo and HP partnering with iFixit to make their devices “repairable”, would you ever consider reviewing a Framework laptop in the future, and if so do you think it would be something you’d consider using as a daily driver?

I think about this all the time.

In some ways, the relative success of Framework in the market parallels Terry Myerson’s original plans for Windows on Arm: He didn’t care if Qualcomm succeeded, the real goal was to finally convince Intel, after years of prodding, to embrace highly efficient mobile hardware and stop pushing only at the high-end of the market. That strategy didn’t work, of course, but thanks to Apple moving to Arm and succeeding beyond anyone’s wildest dreams, Intel finally got the message. And the Core Ultra chipset, with its Arm-like architecture, is the result. It only took a decade

Getting back on topic, Framework isn’t ever going to overtake the big PC makers, of course. But its timing is perfect, coming as it is alongside the regulatory and legal pressure that is forcing these companies to reverse course on the sealed designs of the past and open up their devices not just to repairs by first- and third-parties, but to self-service repairs and upgrades. Framework today is an extreme version of that, but I added a sustainability section to my PC reviews specifically to highlight, when possible, which components in each PC can be easily replaced. And that’s gotten better over time.

So, yes, I do think about reviewing or even using a Framework laptop, in particular a 16-inch version. But I’m also in a unique and weird position where I don’t often buy truly new laptops because I review so many and there’s always something new coming and going. I do buy laptops from time-to-time, but they are usually refurbished or previous-generation models that are deeply discounted (I favor HP, but I just saw that Lenovo is finally starting a refurbished store too), which I use for testing purposes. Anyway, I reached out to Framework but never heard back.

I really like the idea of Framework personally, as I grew up building my own PCs and always wanted a “custom built” laptop, but their prices are a bit too much at the moment especially for those of us not living in the US, so I would love to see other tier-1 OEMs sell barebones or standardized chassis laptops that we can upgrade and repair ourselves.

Yes. I don’t know that we’ll get to that point exactly, I feel like the Framework model is a little extreme for the top-tier PC makers, especially given that whatever profits they can make come from premium models only, and those rely on high-end designs that don’t lend themselves to modular designs. And most customers aren’t asking to replace every single component in a PC (or any components, really). But for those that value this capability, regulations, and laws have forced these companies to make their products more repairable, which is good. And for the make-your-own PC crowd, Framework seems like an incredible choice. I am very curious about this, and I would personally sacrifice a bit of design and portability for it for sure.

(Related to this, I recently asked HP about the possibility of offering buyers of its 16-inch laptops a keyboard choice that would include the numeric keypad that I don’t like and just a straight-up keyboard with no numpad and just some hardware spacers on the side. They seemed very interested in this, though I can’t imagine they hadn’t considered it.)

You didn’t really ask this, but there is also an interesting divide between truly modular hardware (like a PC) and black box/sealed hardware (like a laptop, typically, or a video game console) that is starting to close in various markets. We see that with Framework in the laptop space, which is very interesting. But I wonder about it elsewhere. With Microsoft’s unique approach to Xbox hardware, for example, would a modular Xbox console make sense? That is, you buy an Xbox vNext or whatever, and then three years later you can swap out its graphics card (or whatever) and play the same games but at higher resolutions and refresh rates. Maybe. There was some similar work in the phone space a few years back by Google that ended up failing, but that market is interesting too.

We’ll see. But the move to sealed designs wasn’t just predatory on the part of hardware makers, as there are some benefits to that as well in the form of thinner, lighter designs, something that is particularly important in mobile. It’s nice to have some choice there, and I suspect that if some more modular design in whatever market was successful, competitors would follow the money.

Linux on the desktop

jrzoomer asks:

Set as featured

Report

Paul, is THIS the year of desktop Linux !?

No. 4 percent is still nothing, and most of that gain likely came from Steam Deck, which doesn’t really speak to a mainstream use of Linux as a desktop platform.

That said, Linux is a mature OS, and it’s impressive to me how many companies make distributions and continue to work on improving and simplifying the platform for end users. I tend to favor second-tier Linux distributions like Elementary OS and Zorin OS specifically because they are so focused on good UX and attracting Windows and Mac switchers, but most of the core work to improve Linux is happening higher up the ladder with organizations like Ubuntu, Red Hat/Fedora, Debian, and so on, and there are more mainstream Linux distributions like Mint that are decidedly drama-free and just work as well, and more people are familiar with those.

Whatever your preferences, Linux is the real deal and while it’s not something my wife or kids or non-technical friends would want, let alone understand, we’re not those people. And in an era in which Microsoft has abandoned its customer- and engineering-first approaches to software development on the client (Windows, Office, OneDrive, etc.) I’m not going to apologize for looking around because I don’t like what I see from them. Nor am I going to stop: I’ve been using various Linux distributions in virtual machines on this trip after originally planning to dedicate one of the laptops here to that purpose (I ended up having a weird set of technical issues that required me to keep it on Windows), but I will continue that work on real PCs when I get home.

The thing is, I’ve been trying to replace Windows for decades. And I’ve kept with it because the alternatives—the Mac, Linux, and, more recently, Chrome OS—haven’t raised the bar, and I still very much prefer Windows. The Mac is interesting today because of Apple Silicon, but macOS does nothing for me. Meanwhile, Chrome OS has improved dramatically and probably meets the needs of most mainstream users now, and I am interested in that. And Linux continues to be this mature, stable thing that still has some complexities but is interesting to power users and the more technical; and I am interested in that as well.

Put more simply, I will keep looking at it. But do I see a world in which Linux obtains 10 percent usage share on the PC desktop anytime soon? No. Not really. But then I didn’t see a world in which the Mac would hit 15 percent and will likely achieve 20 percent in the years ahead either. Things change. I won’t ignore that.

Related to this, jeroendegrebber asks:

I have a similar question as jrzoomer: What, in your view, is missing (and thus needed) for Linux to become a daily driver instead of Windows?

This varies by person, but it comes down to apps and services.

One of the things that comes up in Steven Sinofsky’s book Hardcore Software: Inside the Rise and Fall of the PC Revolution is that the Mac’s relative success under Steve Jobs was tied to three factors: Microsoft’s continued promise to keep producing Microsoft Office on that platform, the switch to Intel, and the rise of web apps. There are a lot of problems with Sinofsky, but I am fascinated to see that he confirms the conclusions I came to in Windows Everywhere/my Programming Windows series in We Fought the Web and the Web Won (Premium): Microsoft lost developers for good with the rise of the web in the late 1990s and all of the proprietary, client-side developer platforms it created since weren’t just failures, they were wastes of time that set it and its developer base back forever and left it vulnerable to the rise of mobile with the iPhone.

I mention this because the rise of web apps helps Linux as much as it does the Mac, while Linux also benefits from compatibility with the PCs we already own and is a good fit for older PCs we might still keep around as we upgrade. (This helps Chrome OS with Chrome OS Flex as well.) What’s lacking on Linux is a strong native apps story (which is less and less important each year anyway) and some services niceties, like top-tier support for cloud storage services like Google Drive and Microsoft OneDrive. Which, like other native apps, is less necessary for many now, and can be solved by third-party solutions like Insync anyway.

This is kind of an odd way to look at it maybe, but Linux has been around so long and has never made a dent in the desktop for so long that this longevity may work against it. That is, many people have heard the term, but everyone also seems to “know” that Linux is too esoteric or unfamiliar or unsuccessful, or whatever, to even consider. But as anyone who has tried, or actually uses, Linux knows, it actually works pretty well. The trick is mapping it to the apps, services, and workflows they use. Why someone could switch to the Mac, which I find more unfamiliar and unlike Windows than Linux (or Chrome OS), and be successful is unclear. I hate using the Mac. But surely switching to Linux would be no less difficult. What holds it back?

Looking at my Taskbar as I write this, I see File Explorer, Brave, LibreOffice Writer, Visual Studio Code, Notepad, Notion, Adobe Photoshop Elements, and Paint, and there are other apps I use regularly like Clipchamp, Slack, Zoom, and a few others. Most of those work fine on Linux or have in-box apps that work similarly enough (like a file manager and text editor). Some require a bit of work: I could use GIMP, for example, but I’d have to learn it. That’s doable. I have decades of muscle memory (or, I guess, actual memory) in apps like Paint but, again, doable. Just requires a bit of work. The cloud storage services are a sticking point. But, again, Insync.

Aside from the specifics, which will vary by person, the switch story today is better than it’s ever been. But everyone does have that one thing. If you rely on some Microsoft Office app, switching to the web versions or an alternative may be daunting or a non-starter, for example. It depends.

PC makers embracing Linux would help, I think, but that’s gone nowhere. They barely support Chrome OS too. There’s little doubt they would pounce on any successful alternative. Maybe it’s a chicken and egg issue. Linux could be successful with their support, but they won’t fully support Linux until it’s successful.

I had this question for a while now, but it popped up again after your recent articles about LibreOffice and the reported increase of Linux usage. With the rise of web apps, I think the “app gap” seems to have diminished significantly and learning from my recent use of Linux Mint on my old Surface Pro 4 installing Linux itself and then new native apps is easy enough. So what else is needed?

For you, maybe nothing. For others, maybe too many things. This is a mystery.

Trusting AI

hastin asks:

There’s an interesting article out from Futurism on how the public perceives AI – Trust in AI is down globally from 61 percent in 2019 to just 53 percent, per the Edelman poll. Basically, it says that individuals are becoming more weary of the technology as it gets stuck into everything, and that trust and actionable usage reasons are dropping the general public feel about it. Not to mention the whole “it will replace your job, it will replace artists, it will replace writers” that the industry feels like it keeps pushing.

We are so early in this cycle, and AI is advancing so fast, that this kind of thing isn’t just a slice in time, it’s immaterial. As AI changes and improves, so too will our opinions of it. Honestly, this is like anything else in life: It’s easy to be ignorant about something you’ve never experienced, but once you experience something you open your mind to it. That people still doubt AI is “real” or whatever is pure ignorance at this point, as it’s easy to experience it now, and it’s real. But not trusting AI? That’s just human nature. There were people who doubted the steam train, the automobile, the airplane, and whatever else. This kind of thinking was effectively lampooned on the show 30 Rock in its “Beeper King” story lines.

Personally, for me, I think too many things are getting branded as “AI” and muddying the water. The photo editing improvements, automatic summaries, and other features would have happily just been considered part of the “magic” of a tool, but are now labeled [as AI].

This is no different than any other tech wave. The Dot Com boom was about how the Internet was going to raise every ship, but the resulting crash simply level-set the industry, and it’s not like we stopped using the Internet. It didn’t just grow, it’s all-encompassing, an essential service that we all expect and rely on. AI is like that. Yes, we’re going to go through a hyper-growth phase in which there will be some marketing overreach, and, yes, there will be consolidation as some of the least-trustworthy or useful products and services will come and go and confuse matters on the way. But the truly excellent AI products and services are obvious. And they are not going away. They are the foundation for what computing is going forward.

I often explain that I used to use computers for hours on end in the pre-Internet era, but aside from some obvious things, like playing games, I have a difficult time remembering what it was I was doing, exactly. AI will have the same effect: We won’t be able to imagine the unsophisticated world from pre-AI or how anyone got anything done then.

Do you think tech companies are leaning too hard into the words of “AI”, and not the actions of what AI can actually do for people? How do you feel the public will see this technology in 5 years?

As essential and obvious. Marketing isn’t going to kill AI. Nothing is.

Then again …

The dark side of AI

JustMe asks:

Starting with AI, its use and misuse – there is no denying the AI train has left the station. While it can be incredible (helping to read scans for physicians, reading ancient burnt scrolls), it can also be scary – the BBC recently reported on pictures that were generated of a politician at a party and on some house steps with a particular voter demographic the campaign is after. The scenes depicted never happened. These pictures were NOT generated by a political campaign but rather supporters. One of the pictures authors said bluntly that he never claimed the pictures to be real, he saw them as almost a political cartoon, and if your vote was swayed by a Facebook post it said more about you than him. This year in particular has many elections worldwide. The question I have is how do we police such behavior? Could the tech companies have any image generated by AI somehow watermarked so that anyone viewing the picture was aware of what they were seeing? Barring legislation, is there anything we CAN do? As Americans, are we stuck cheering on the EU?

This is the trick with AI and is perhaps an excellent rejoinder to anyone who feels that AI isn’t “real” or whatever. No, it’s a significant advance. It’s happening quickly. The law has never been able to keep up with tech advances, but AI is happening almost exponentially quicker than any previous advance. There will be problems. There are problems.

That said, we’ve been dealing with photo trickery ever since we’ve had photos. This notion that “photos don’t lie” has always been nonsense, just like “seeing is believing.” We are easily tricked. AI is making it even easier. But we’ve had deepfakes for years.

What’s needed, of course, is regulation. In the sense that tech has been under-regulated if not unregulated for years, AI drives home the need like nothing else. Right now, the EU is leading the push to rein in Big Tech, but we’ve seen action elsewhere, including the U.S. which is set to launch a major antitrust charge against industry darling Apple any time now, among other things. So don’t give up hope: The aggressively powerful nature of AI will scare regulators and governments straight quickly. The issue is the short-term: In the same way that Facebook helped undermine the U.S. election in 2020, AI will have a major impact this year, one that is far more dangerous than anything Zuckerberg could dream up.

My worry with AI and regulation (including self-regulation) is that we’ll follow the TSA model where we are always responding to a previous incident and not proactively doing the right thing. Regulation of AI, and a standard way of identifying AI “fakes” (really, any AI-generated content), isn’t just necessary. It’s necessary now. But that’s not how the world works, sadly.

More on AI – as a content creator, what steps can you take at this point to protect your work? You posted recently about AI scraping your stuff and not liking it – could you bar the AI bots with something like a robots.txt file to protect your site? If you do that, what impact (if any) does that have on legitimate searches? What else could you do?

Honestly, there’s nothing I can do. There are supposedly site protections I could put in place, but from what I can see, these things work like “Do not track” in a browser, meaning that they are simply ignored. This falls under the regulation thing above, but it’s already too late. Everything I’ve ever written is likely out in the world, feeding AI already. You can’t close the barn door now, the horses are long gone. And replicated across multiple AIs, I bet.

Amazon and Android

helix2301 asks:

Paul do you think one of the reasons Microsoft and Amazon are getting out of Android is because a while back amazon announced they are going to their version of Linux on all fire hardware and using meta-framework and doing away with Android. They were Microsofts biggest Android store could this have something to do with this subsystem being discontinued?

Amazon didn’t announce that they were abandoning Android, there were news reports that Amazon was working toward creating its own OS for embedded devices like FireTV. I suspect this is like what Google is doing with Fuscia (and, soon, with bringing Wear OS to RISC-V), where something even more lightweight is needed. I don’t see Amazon jettisoning Android for its Kindle Fire tablets and the like.

We can speculate about why the Windows Subsystem for Android (WSA) failed. My guess is that it came down to Microsoft focusing its developer story on AI, and that WSA was simply too unsuccessful as a developer product, one with no AI story, to matter.

A Microsoft executive named Andrew Clinick—who, in a neat bit of coincidence, previously represented Microsoft’s last-gasp client developer platforms, UWP and the Windows App SDK (Reunion)—offered up a similar view on Twitter this past week. His current job description is a bit wordy, but here goes: He’s a product management leader of Microsoft’s platform development and strategy, AI developer platform. That is, he has nothing to do with WSA, and never did. But he provided some insider information that could point to part of the problem with WSA.

“WSA was an amazing project to work on but doesn’t matter how cool the tech is if you don’t have apps in the store,” he tweeted. “Turns out standing up a store and attracting apps is mighty difficult. Then you factor in that Google Play services isn’t there. Difficult to find an app with a large audience that doesn’t use Google Play services.”

“Store revenue pays the bills and salaries of the WSA team,” he added. “WSL accrues to Azure and Visual Studio so has a profit center.”

In other words, WSA cost Microsoft money to run and it was never going to be profitable without apps, and it never got decent apps with just Amazon. Google was never going to partner with Microsoft on WSA, so it was never going to make sense. Amazon was always the wrong partner there.

Windows Media Audio

helix2301 asks:

My question is I have some older music in .WMA format I converted it to Mp3. Is wma format still a thing? I remember ripping music using windows media player was very good at that.

No, not for a long time. WMA is a classic Microsoft fail in that it hasn’t been updated in decades and isn’t used anywhere. Today, the standard compressed audio format is Advanced Audio Coding (AAC), which is to audio what MP4 is to video.

Hot-air ballooning

VMax asks:

Super random one – after watching the videos you posted of the balloon flight, the professional one was obviously taken using a drone of some description. Was the noise troublesome? One of the best things about balloons (IMO of course!) is the silence and stillness that you experience, so I wonder how the annoying drone noise vs great drone footage balance works out.

No, there was no real noise from the drones (there were several), which rarely got closer than 15 feet to the balloons. But hot-air balloons can be quite loud: Each time the pilot ignites the gas to create the flame that pushes the balloon higher, there is a very loud burst of sound. It freaks some people out.

Related – is this a standard thing there, or was it some kind of balloon festival? Looked like an amazing experience either way, I’ve never flown with more than a single balloon and would love to be amongst that many!

This happens there every single day. And this wasn’t even a weekend or whatever, we went on a Monday. I’ve done this three times near Mexico City now (my wife has gone four times), and we’ve flown in hot-air balloons previously in Albuquerque, New Mexico and Phoenix, Arizona. But aside from the annual Albuquerque Balloon Fiesta, which is amazing, I’ve never seen so many balloons in the sky at once as occurs every day outside of Mexico City.

The balloon videos we posted on this trip are:

• Pro footage: Balloons over Teotihuacán 2024
• Balloons over Teotihuacán 2024
• So many balloons!
• Bumpy landing
• Balloon ping-pong

The enshittification of Windows

JustMe asks:

Windows 11 – my personal daily driver runs Fedora 39 these days. I do have a windows machine for those times I really need windows. I undertook to upgrade from 10 to 11 recently, and was almost taken aback by how much Windows gets in your face and how busy the UI is these days. Many of the notifications of course can be turned off or muted, but the difference between Fedora and Windows 11 seems stark to me. When I’m running Fedora, I hardly notice the OS. When I’m running Windows, I know it. I dont hate W11, I just wish it would pipe down, and I’d really like more control. I wish it would respect my choices – I am looking particularly at a group policy setting to get rid of the recommended area of the start menu that the Pro version of the OS apparently ignores. Oh, and for the love of all that is holy – Microsoft, will you PLEASE do something about File Explorer in Windows 11 stealing focus. Trying to install recently downloaded software got annoying quickly. Not really a question there, more an observation – going between Windows and Fedora, why does Microsoft insist on getting in your face – the OS itself, Teams, Office…cant Microsoft just calm down?

I feel like I’m the canary in the coal mine when it comes to Microsoft enshittifying Windows, and while I find it astonishing that the company would treat this iconic product and its own customers so shabbily, I don’t see any end to it. Windows has been forced to conform to conditions in which it cannot excel, first, the cloud-centric era instituted by Satya Nadella and more recently by AI.

But the underlying theme here is the same: Microsoft first shifted from selling boxed copies of software to an annuity model in which it is paid monthly in perpetuity decades ago, but because Windows is most frequently obtained with a new PC, and PCs last much longer these days, it just can’t exist well in this system. So Microsoft has tried various ways to get Windows users to pay more, and pay more often, via various subscriptions that mimic, in some ways, the enterprise agreements it has long had with companies. It’s not working.

That’s when enshittification happens: It’s this perfect storm of a product used by a billion people who all paid for the thing once and then stopped paying. Microsoft has seen great success elsewhere in the company getting companies and individuals to keep paying for the things they use, but not Windows. And so the strategies have gotten more aggressive. It sucks. But it’s the world we live in. This is what makes other platforms so appealing: Linux, Chrome OS, and macOS do not do this.

I mentioned above that I’ve been trying to replace Windows for decades, which is a cute way of saying that what I’ve really been doing is looking closely at its alternatives for a long time. But it’s not cute that Microsoft is now doing everything it can to drive me—and, I think, others—off Windows too. It’s not good.

Working while away

JustMe asks:

Travel – do you notice any difference between working in Mexico and working in the US? Are you more productive one place than the other?

I’m semi-unique in that I’ve been working from home for 30 years now, and starting in the early 2000s, we would spend at least a month outside of the U.S. each year, and so I’ve also spent a lot of time honing what it looks like to work during those trips. By the time we got to Mexico, I was already very accustomed to what I think of as a More Mobile setup, but actually having a second place (here or anywhere else, I guess) is, if anything, easier because I can leave equipment here. And as of this trip, we both have everything we need here in Mexico. We could just stay here forever from a work perspective and it would be fine.

We have broadband Internet here that’s as fast and stable as what we have at home. We have phone plans and whatever, and our phones work normally. There are specific little issues around some services not being here or whatever, but nothing major. (I wanted to temporarily subscribe to Clipchamp Premium so I could create a few 4K videos for example, but I couldn’t do that from Mexico because the app and my account were from different locales. I used a VPN to make it work, not a big deal.) But day to day? It’s just like being at home. I have a desk with computer stuff, connectivity, a phone, and I have laptops so I can move around the apartment as needed.

I often feel like I could go away, never tell anyone, just work normally, and no one would ever know. The only obvious thing is that it looks different behind me in podcasts or work calls.

Backups

JustMe asks:

Backups – part of my upgrade was of course to make sure my data was backed up. This gave me pause for thought, though – what is your current backup methodology? In my case, I typically back up the data -docs, pictures, data, etc so that I can easily just reset the machine, but do you have a particular piece of software you use? What do you think of the (new to me) W11 backup app that backs things up to OneDrive? What about the old school tool that is still in Control Panel which images your drive?

I wrote about this in Roll Your Own Windows Time Machine (Premium), but you might think of it in terms of this More Mobile setup as well. Traditional backup is obsolete and unnecessary in a world in which you can sync your data to the cloud and to other PCs instantaneously, reset/install the OS in minutes, and easily automate the installation of apps. (The missing piece is automating the configuration of those apps. I’m working on that, but will point out that using web apps solves that problem in a blow to the PWA deniers because you configure the app once in the cloud and it’s always configured properly.)

The new Windows 11 Backup is a joke and can easily be ignored. The old tools in Windows—classic Windows Backup and File History—are obsolete and should be avoided.

Why ask why?

JustMe asks:

Why is it Microsoft will not let you delete an XBox gamertag? I undestand you might need it playing online – but as a PC gamer who plays almost exclusively in single-player mode, I dont get it. The culprit in my case was Minecraft. Prior to my OS upgrade, I got to looking at my Microsoft account settings. Lo and behold, I had a gamertag that was set pretty much to default settings – soley for Minecraft, which I touch once an eon. Even had a ‘follower’ somehow. Set all my settings to private, and then thought to just delete the gamertag as I wont ever use it online. Nope. You can transfer it, once, but Microsoft wont let you delete it. Why?

“Why” questions are tough because the reasons behind such a thing are rarely obvious and, in this case, are likely tied to the product’s legacy. My guess is that this is tied to how Gamertags were implemented back when there was just a single Xbox console and a new service called Xbox Live, and that in being subsumed into what we now call a Microsoft account, which is an entirely different codebase from an entirely different team, that things just got complex. Microsoft still doesn’t offer a way to merge Microsoft accounts, for example. There are lots of things like this.

Basically, I have no idea. Microsoft is confounding. But I have to think you could contact Xbox support and ask them to delete it for you.

Windows 12

madthinus asks:

Is there a Windows 12 cooking or are we once again on the “final version of Windows”?

Microsoft never claimed as an organization that any version of Windows was the final version, but even Windows 11 is not “a version,” it’s been three versions of Windows so far. We should just think of it as a brand. (He says, trying to support a “Windows 11” book like an idiot.) Anyway, the consensus is that what might have been called Windows 12 is now called Windows 11 version 24H2, and while we may see something called Windows 12 in the future, we’re sticking with the Windows 11 brand for now. It’s a rare burst of clarity for a part of the company that doesn’t even understand the term. Meaning that now that I’ve written this, Microsoft will of course announce Windows 12.

Mobile

Akis asks:

Hey Paul. Do you think there’s room for a 3rs player in mobile, apart from Apple and Google?

No. To date, the industry has been quite clear that two major mobile platforms is all that anyone wants or needs. That said, as regulators finally crack down on Big Tech and these platforms start to open up more, we could at least see a new kind of heterogeneous mobile world in which the strict alignment of OSes, apps platforms, and stores are no longer enforced. An iPhone that doesn’t require web browser makers to use Safari is obvious enough, but Apple allowing iPhone users to actually remove Safari is rather incredible. It kind of opens up things to new, unexpected possibilities even within the confines of these two platforms.

This includes, I think, the thing that Apple has fought so hard to prevent: Web apps as a first-class platform everywhere, not just in most places. As that happens, and it will, the underlying platform becomes less important, just as we’ve seen in PCs. And then it’s possible, unlikely, but possible, that other mobile platforms could be successful. But they won’t be about native apps. And that’s a very different world.

We’ll see.