This Week in Tech 575
Leo Laporte: It's time for TWiT, This Week in Tech. We have a big panel for our very last TWiT at brickhouse studios. Join Greg Ferro from The Packet Pushers Network, Allyn Malventano from PC Perspective, from Engadget, Devindra Hardawar, and Alex Wilhelm from Mattermark as we talk about the latest news in the tech world, including the Galaxy Note 7, what 5G really means. It's going to be an amazing panel, you stay here. TWiT is next.
NETCASTS YOU LOVE FROM PEOPLE YOU TRUST. THIS IS TWIT! Bandwidth for This Week in Tech is provided by CacheFly at cachefly.com.
Leo: This is TWiT, This Week in Tech, episode 575, recorded Sunday, August 14, 2016.
This Week in Tech is brought to you by Casper, an online retailer of premium mattresses for a fraction of the price, because everyone deserves a great night's sleep! Get $50 off any mattress purchase by visiting casper.com/TWIT. Enter the promo code TWIT.
And by Audible.com. Sign up for the platinum plan and get two free books. Go to audible.com/twit2.
And by SquareSpace: SquareSpace is the best way to create a beautiful website, blog, or online store for you and your ideas. Go to squarespace.com, enter the offer code TWIT at checkout to get 10% off.
And by GoToMeeting: Be a meeting MVP. GoToMeeting is the online meeting tool that lets you easily collaborate and present from anywhere with one click. So you'll always put on your best performance. Step up your meeting game and start your free 30-day trial at GoToMeeting.com.
It's time for TWiT, This Week in Tech! The last TWiT from our beautiful brick house studio, so we thought we'd gather first of all some of my favorite people and then have them all in studio at the same time. I think this is going to make a fun TWiT. I think it's the first time in many years we've had five people all seated together in studio. It's great to have you. Let's start over here with Alex Wilhelm in his jet T Shirt. Instant memorabilia.
Alex Wilhelm: That's why I wore it. I thought I would wear someone's uniform for a topical thing.
Leo: Wall Mart just called. They're offering you a buck and a half. Will you take it?
Alex: Plus stock options? I'm in. 50 million dollars in losses, I'll sell to anyone who is buying.
Leo: You have to think Mark is going wooh. That was close.
Alex: No one knows.
Leo: By the way, Mattermark. I should give you a plug. Your employer, mattermark.com. Also here from Engadget, senior editor, Devindra Hardawar. Always welcome.
Devindra Hardawar: Thanks for having me.
Leo: You've never been in studio before.
Devindra: No. Never. I'm based in Brooklyn, so I'm always calling in, but I'm here for Intel conference. So, perfect timing!
Leo: Yay! I think a lot of people are here for IBM.
Devindra: Yep. It was hard to get a hotel.
Greg Ferro: Same for us.
Leo: Greg Ferro here from the Packetpushers Network. packetpushers.net. Greg, we've had you on the show before, but this is the first time in the States. You're based normally in...?
Greg: Chiltman in the UK. In the Midlands.
Leo: Do you have moors there?
Leo: What is a moor?
Greg: A moor is in the Scottish highlands, and it's normally rolling hills covered in heather.
Leo: I think of the Hound of the Baskervilles in the moors and I never knew... they were always going out to the moors.
Greg: Down near Devon and Cornwall. I think it's actually a geographical feature. I'm in the midlands, the rolling hills and the rock walls and the cows.
Leo: Beautiful. A lot of call for networking out there.
Greg: It's a long story. I ended up there after marrying an English girl.
Leo: Because you're originally from Australia.
Greg: I am. I'm Australian, yes. I left ten years ago, but all my work is now in the US.
Leo: Are you here for IDF?
Greg: I am here for IDF and a bunch of other things, yes.
Leo: I'm really thrilled to have Allyn Malventano here. He writes for PC Perspective, former "don't get excited guys," NSA agent. He works in the US Navy for many years as a Mariner, you were also an analyst at the national secrecy administration. Now he's the guru of the hard drive. I've been friends with Allyn for almost ten years. I remember calling on you constantly. The Ultimate Gaming machine. You had the craziest setup. How many SSDs did you have?
Allyn Malventano: When? It changes all the time.
Leo: You had that? I thought you had hundreds of them all ganged together.
Allyn: I was pitching it to you, because you were talking about velociraptors at the time.
Leo: Ten thousand RPM spinning discs...
Allyn: I found that... I forgot who made it, but you put six SD cards in it, and it raided them.
Leo: It gave you 40 gigs. We've come a long way, baby. You were telling me about SSD that is 60 terrabytes now?
Leo: That's what you're here, the flash memory. We'll talk more about that. Wow. What a great panel. Nice to have you all. Thank you all for being here. You are here. The next TWiT we do will be for better or for worse... we may not have lights. We may not have electricity, but we will be at the new studio, the east side studio. The way this is going to work a week from today, I'll do the last show from our studio, the radio show on Sunday afternoon, and we're all going to board, we've rented a cable car. We're all going to get on that, including a camera with a live view attached, and we'll all make the transition over to the new studio, and people will join us. If you're here next week, you'll be asked to carry a box.
Alex: Worst pitch ever. Carry this back and forth.
Leo: One box is all I ask.
Alex: Free pizza and beer.
Leo: That was a seminal moment in my growing up. I knew I was an adult when I said I'm going to help you move. Then you know. I was in my 40s. I am an adult now, I will not help you move, so do not ask.
Greg: Like a shopping cart, actually move.
Leo: I did that in college.
Greg: Actually it was a squat. I was moving from one squat to the other.
Leo: Do you want to tell us more about your early days? Interesting.
Greg: This isn't the forum. Maybe if there was cocktails.
Leo: Where should we begin? This week Google let it out that they have a new operating system that they're working on. It's Fuchsia. Fuchsia is the codename. It's open source, they posted the code on Github of all things. Is it a IOT operating system? Maybe, but maybe not. It looks like it has desktop capabilities as well. It is not a Linnux Kernel. Fuchsia is based on something called the magenta Kernel. That's based on something you might know if you pay attention to this stuff called the little kernel. I don't know if it's considered a real time operating system. It is SSMP aware. You can use it with a desktop processor. It's not small in power. It's small but mighty. Why would Google want a new operating system? What might they be using it for. Greg?
Greg: I think what you're seeing it for is this broad trend across all of technology around convergence. That which was separated is now converged. This is a trend that Apple has led with iPhone. You get your operating system and your hardware together. Google is doing the same thing with Chrome. You get your device, the operating system is pre-installed. The years of separating is passing us by. Apple watches IOT and the operating system builds together.
Leo: But this sounds like the opposite. This sounds like fragmentation.
Greg: It is. It's stripping out all the stuff that you don't need.
Leo: This would be eventually the operating system.
Alex: Is this similar to the Windows Everywhere strategy Microsoft is currently working on? This feels like a pan OS thing. Android is terrible when it's not this size.
Greg: The way operating systems are structured today is there's a lot of functionality in the operating system. It's not just running the CPU, booting the hard drive, getting the operating system running. What we're seeing is thousands of functions built into the kernel. Linux, for example. If you look at other operating systems like Kinex, these are micro kernels that have the bare minimum functionality, and everything else is...
Leo: Apple uses a micro kernel called "Mock." Same idea.
Alex: This is not using that same structure.
Leo: Linux is not. Though it's funny. Linux started as a more modular kernel that you would plug in what was needed. As it ships today, unless you build your own kernel, it's likely that you'll have support for every possible kind of...
Greg: The person who is behind this has developed many kernels before. I see this as a reversion towards micro kernels. Back to real time operating system with all the functionality stripped out so that you're maximizing your battery power, there's no unnecessary apps, and you have to build a modular packaging system on top of that. If you want printer driver support, you add printer driver support. It doesn't come as a standard feature.
Leo: An unsophisticated user would look at this and say they don't want to do Android anymore because it has the features, in fact it's got the material design interface. They're using something called Flutter, which will give the material design. It's really designed around using as little battery as possible. It's almost a real time operating system. It feels like a phone operating system, more than a desktop computer.
Greg: I think everybody points the finger at Google and says, "This is Chrome. Why is the Chrome operating system different from the Android operating system? Why don't you converge these two together?"
Leo: So this is that.
Greg: I would speculate.
Leo: They're putting Android on Chrome OS. You can get the Android store on Chrome OS, and they're running Android and a sub system on Chrome OS.
Greg: They're running away from Oracle. The Oracle free Java suit versus Google. Android operating system is all built in Java and this is all compiled. It might be a backdoor, a way of saying if this comes unstuck.
Leo: More than that. I think that they're running into... I think the EU, and Russia just fined them 6.8 million dollars, a laughable amount. I'm going to fine you a penny. But the EU has been down on them because of the way they manage Android, and it's funny because in a way they're getting into trouble because Android has an open source arm as well as a proprietary arm. They're not beating on Apple for doing, for requiring that all phones have Apple IOS on them. They're mad at Google though because they're making some restrictions on what people want to use Android for.
Alex: Android's global market share is far higher than Apple's.
Leo: Make's more of a difference.
Alex: Right. We're very accustomed in America to see a lot of iPhones come on Parody, but outside of this country, that's not the norm. So to make the antitrust concerns a lot more relevant than the Apple comparison...
Leo: I'm not saying... well I am saying. But this would be Google's attempt to say, "Fine." maybe a do over.
Greg: What about security? How many vulnerabilities are there in Android that we talk about every week? There is a massive insecurity in Android. Something like 40% of Android phones are compromised today.
Alex: Is that due to the cycle for most consumer handsets?
Leo: That's more the problem. Google is fixing it, but they're fixing it in the mainline.
Greg: It's difficult to update the current Android system with the current architecture. You have to replace the entire operating system. If you had a micro kernel you can start to marginalize the patches...
Alex: Can I confess? I don't know what Micro kernel means.
Leo: It's a kernel that's really small.
Alex: I understand. Pretend I'm a normal consumer.
Leo: It's like risk versus sis.
Alex: Ah. That makes perfect sense. Great.
Greg: What you're doing is reducing the surface of the kernel. The Kernel is the core of every operating system. Your Linux Kernel is a fairly substantial piece of work and it's been evolved over 20 years and has hundreds if not thousands of functions built into it. Every one of those functions has to work perfectly for it to be stable and reliable. When you upgrade it, you replace all of that code. with a Micro kernel what you do is pull out all the stuff into Surrounding pieces. Into modules, if you will. You decouple them from the kernel. The Kernel now has the bare minimum of attack surface. Least amount of code, a minimal amount of functionality.
Alex: I want to turn that answer into a dance remix. That was incredible. It had a good flow to it.
Leo: Just ask the chat room. It's interesting to read. I'm reading an Android police article that because they made it open source, people are kind of working through the code. There's interesting things to see there. A permissions based or capabilities based security model, their obviously paying attention to that. I find it interesting they have a renderer in there which supports open GL or Vulcan. Many of the features of Open GL or Vulcan. I guess some of it could be for UI, but it could be for gaming too, right?
Allyn: Vulcan has a more efficient API.
Leo: Uses less battery. Less CVU.
Greg: That's what Apple's metal does. Because they bind the metal very closely, the reason it's called metal is it runs directly on the Silicon.
Leo: It's got Direct X, which is kind fo that idea as well, was it not?
Allyn: Kind of, but there was more overhead.
Leo: Also built in support for 32 bit, 64 bit arm, and 64 bit I86.
Greg: Intel has a number of power efficient atoms that they're trying to drive into the market. Intel you have Flash, or the year before they were announcing a smartphone. A Prototype that was smaller and thinner than an Apple iPhone. So the Apple iPhone was 6.1 MM and Intel had 5.9 milimeter prototype. It hasn't really taken off because the process is still advanced ahead of what Intel is doing. I think what Google is doing here is giving Intel a chance to come back into the game and get their chips into Smartphones.
Leo: Somebody in the chatroom mentioned this is somewhat like what Samsung is trying to do with their own operating system. Be less dependent on Google, less dependent on Android. Also, by the way, according to Google, this will be available for the Raspberry pie three as well. Interesting.
Devindra: That makes sense for the lower powered computers. I wonder what this means for IOT devices. It may be, other things along those lines. If Android is too heavy to run on some phones these days, as we get smaller and smaller devices, maybe if IG networks that aren't using much data speeds, we need something to run on.
Greg: We need IOT devices that run for ten years.
Alex: As discussed earlier, but notably Microsoft and Google are working on this problem, but Apple isn't.
Leo: OS 10 was a micro kernel. However, it's a pretty old micro kernel.
Greg: It is. But Apple continues to iterate. They've iterated the kernel substantially.
Leo: IOS and watch probably run on Mock as well.
Greg: I'm not too sure, but their ability to design the CPU, to take advantage of the kernel and to squeeze out the power show you the power of a micro kernel. Also the ability of an Apple phone to download its patches reliably. Updating your Android isn't necessarily a foolproof game.
Leo: But that also has as much to do with the economics of it and Apple is the only manufacturer of the iPhone. They've strong armed the carriers in the US to say, "You're going to push for updates and if they're available..." Apple used to push big updates. They're now down to these delta updates. It took them a while to get there.
Greg: They still have people claiming they updated their Android and it crashed and burned.
Leo: Never hear that on the Apple side.
Devindra: It's a little more convenient now. You can say download the update in the middle of the night, and it will install this for me. It's nice.
Greg: If it was a really bit problem, you'd hear the websites complaining that Apple is blowing it, and Steve Jobs would never have stood for it.
Leo: We hear that all the time. Maybe you don't hear that in the UK.
Devindra: I wonder if this is trying to fix some of the original problems in Android. You were talking about being so reliant on Java, but also the idea that maybe they didn't expect Android to get as big as it did, and now it's everywhere and it's like Oh My God. This is the wrong engine for what we're trying to do right now.
Leo: Didn't plan this, obviously. It's pretty clear that they had no idea what was going to happen with Android. It's funny. One of the reasons I'm fascinated by this is we had this conversation some months ago. Where is the next operating system coming from? My conclusion, our conclusion in general was, "Who cares?" It's no longer about the operating system. The cloud meant at this point that we're operating system agnostic. All data, even many of our programs... Oh. That's my phone talking. Just ignore that. My daughter calls right in the middle of TWiT every week.
Alex: That's great. If you're a kid and don't want your parents to actually answer the phone but say you've called. Call when they're at church, or... Sorry to blow up our spot.
Leo: Dad, you're working now? I'm so sorry. Oh.
Alex: By the way, did you send the check?
Leo: That's the other thing. The message will say, "I'm out of money."
Greg: I think you're right. Most people in the Cloud, I think there were statistical surveys showing 70% of the people running Ubuntu in the Cloud.
Leo: Linux won a long time ago. Android is Linux. Chromebook is Linux. Look how well Chromebooks are selling all of a sudden. They're huge. In fact a great article in Wired Magazine this week from a writer who said, "To students, your next laptop for going to college should not be Windows or Mac, it should be a chrome book." I agree with him. Especially in light of the fact that the Android Play store is coming to Chromebooks.
Alex: Why would you ever buy a Chromebook over a fully featured OS 10 computer?
Leo: Why wouldn't you?
Alex: Because it doesn't do... it only does one thing. Chrome OS.
Devindra: It doesn't have the expandability of many apps.
Leo: that's exactly why you buy it. It doesn't have the complexity... it doesn't have the security issues.
Alex: What if I have to run software from my school? My University...
Leo: What software are you talking about?
Alex: Math lab.
Leo: If you're a physics or a math major...
Alex: Or computer science. Or design. Or Philosophy.
Greg: I've worked with a team of schools in the UK, and there's 20 or 30 private schools in the chain. They're switching to Chromebook for two reasons. First of all, a Chromebook is desposeable. If somebody drops it, you get another one. The price point is fair. What you can't use is an Apple iPad if you drop it and break it, you're blowing up a substantial amount of money.
Leo: I would guess that there will be if there aren't already, web based interfaces. I'm sure there is. Statistics programming language. But you could easily do cloud based coding. That's not a problem. There are plenty of solutions like Code anywhere that let you do that.
Devindra: It's easy to see why Chromebooks have taken off in schools just like that. I used to work in NIT for a long time. The idea of having a cheap, disposable laptop you can set up easily if it breaks no big deal, that's great for institutional purposes. For a student, for something that they're going to bring and rely on for their entire school life, you want something a little more...
Greg: Most education systems from kindergarten to year ten don't need any of that stuff. The advantage of Google Chrome is everything exists in the Web.
Devindra: IF you're going to college, you're preparing to enter the workforce, and you need to run the apps you're going to be running at work. You're not going to go to you know, a video editing lab.
Leo: Here is Matt Lab on the web.
Devindra: Most of those will exist on the web and get there somehow, but for big things like editing multi media files.
Greg: That's a fractional usage. In the UK, something like 80% of real estate agents use Google Docs and they use ChromeBooks.
Leo: I bet you they have a lot of college kids these days editing videos on their Smartphone, not on their desktop computer. You have a multi media device already, it just happens to be in your pocket. I think part of the problem is a lot of us, even younger people like you guys, still think of a computer as a desktop or a laptop device. I think if you're in college today... maybe that's not...
Greg: You can edit a video in YouTube itself. I wouldn't want to be editing something like this in a YouTube creator, but...
Alex: If you only need surface level locations, you can get away with web based alternatives that are...
Leo: It's my position that most people, even college students, that's what they need.
Alex: I agree with you entirely.
Leo: Nobody watching this show... we also have the expertise to stay out of trouble to fix a problem. I think those people should be using frankly, Linux.
Alex: That guy says I use an Atticus, which is a great rejoinder to this entire conversation of micro kernels.
Greg: People watching this show and certainly around the table can afford 2,000, 3,000 dollars for a computer and justify it.
Leo: It's an important priority for us.
Greg: If you're a school and you're buying ten thousand computers, it's a big difference between a 300 dollar Chromebook and a 600 dollar iPad.
Devindra: I take bigger offense as I'm seeing Chromebooks get more and more expensive. HP just released one.
Leo: I don't mind seeing a little bit more expensive ones that are better quality. For a lot of Chrome it's netbooks, right? They're just crap. But a 400 dollar, 500 dollar aluminum like the Acer or the dell, those aren't... that's still. The point is... Nobody should buy a Pixel.
Devindra: Going past 500 is kind of crazy. Some of the new ones are going higher and higher.
Alex: Is 500 in the upper threshold of what you should pay for a Chromebook?
Devindra: that's what it sounds like.
Leo: You should get a touchscreen now because of Android. You should get 4 gigs of Ram. I don't think... maybe eight. You don't need a lot of storage in a Chromebook, obviously. I don't think you need an i5. An atom of a...
Alex: But when I run more than 15 tabs in Chrome it's going to slow people down. That to me is...
Leo: This is bad usage. I don't know why so many... it's always people on this show, have thousands of tabs and bitch because it's all so slow. Why the hell do you have so many tabs open?!?
Alex: I kvetch, I don't bitch.
Leo: Close a freaking tab once in a while!
Alex: I will not, sir, close any freaking tabs or fracking tabs, because I'm using them, darnit.
Leo: You're not using 15 tabs.
Alex: You're working on three different articles, you've got a bunch of tabs open, browser windows, I need a computer to do work.
Devindra: Some college students, sure. If you don't need much going into college, that's fine. A lot of kids, you may not know what you want to start getting into. Having something that locks you in, within a year or two, just like you change your major, you may need to change your computer if it turns out you need.
Greg: When I was 18, I should have been buying a family Sudan ready for when I have kids?
Leo: Get a van.
Greg: When you're 18, you buy a stupid car, you turn the wheels inside out.
Leo: That's why you buy a Chromebook. Buy a cheap computer, because you don't know what you want to do with it.
Greg: Then if you do want more, you spend big, and buy the BMW three series.
Alex: This analogy has died, guys.
Leo: How many tabs do you have open right now?
Greg: My home computer has over 80. I have a special app that tracks all the apps so I can save the tabs.
Leo: The reason you don't close it is because it takes a while to reload?
Alex: Generally because I'm researching around 8 to ten topics on a different thing.
Leo: I guess I just don't work that way. Steve Gibson is all about tabs.
Alex: I have 40 across two browsers.
Leo: How big are the titles?
Alex: They're little. I'm not trying to be negative or contrary, I just think if you're going to do more than a very simple task at one at a time, having an efficient computer is my preference and everyone should have the ability to have...
Leo: I'm not telling you what to get. You should get whatever you want. Anybody who wants to take on the complexity, the security issues, the responsibility of a general purpose computer absolutely should. I wouldn't recommend you build your own computer either! It's a great thing if you want to, but you should know what you're getting into. I feel, unfortunately, a lot of people are doing what we all did ten years ago, because you didn't have a choice, which is buying Windows or Mac. It's now inappropriate. It's time to start thinking either of the tablet or the Chromebook.
Alex: It's interesting when you reach that threshold. Security on Windows Devices is now inherent and easy to use. During the Windows XP era, the ten years ago you're describing, that's when security was the worst.
Leo: Here's an interesting question. I had a great caller on the radio show, lovely lady, Rose. She is a community manager for a synagogue, and as a result sadly to say, a target. Her Facebook account, which I couldn't understand at first. I spent a half hour with her on the air. Her Facebook account, unaccountably, people were posting Turkish language spam links to porn sites on. All her friends had unfriended her. Her synagogue blocked her posts, Facebook had warned her she's spinning out too many links. Clearly somebody had co-opted her Facebook account. I asked her all the questions you had asked. Did you change your password? yeah, I change it all the time. I encouraged her to change it to two factors. I said make sure you disconnect all your connected apps. I thought maybe some app is posting on her behalf. She said I did. I said, you've checked all your browser plugins? No.. what browser? She's on a Mac using Chrome.
Allyn: She checked almost every box.
Leo: She did. She checked almost every box. You were listening. Finally we found out she had a man in the middle attack going. Not from a bad guy, but from her anti-virus. She was running a vast anti virus. I asked her to look at the Facebook certificate and it was going through a vast. My theory, I can't confirm it because I can't see the computer. It wasn't a Vast doing this, but that the Vast had somehow been hacked. The certificate was not to Facebook it was to Avast.
Greg: And they were compromised. They were binding route certificates. Their private route certificate was lost some time ago.
Leo: That's right! I remember we were talking about this on Security Now. So they had been compromised. There you go.
Greg: Your browser has a certificate score in which you can obtain the route certificates. When you install Avast, it puts its own certificate into the route store.
Leo: A lot of anti-virus software do this mis-guidedly. Lenovo did it with Superfish.
Greg: The problem here is Facebook. When the client presents a certificate to Facebook and they see it's not the user's certificate, they see an Avast, they should see that it's been a man in the middle.
Leo: However, it's a quasi-legitimate man in the middle that this anti-virus was using to protect her.
Greg: What those viruses are often doing is capturing the traffic that goes through and selling it for money. that's what Avast is doing, and so is Verizon and Comcast and AT&T. So as your data crosses their networks, they're capturing that data and selling it off.
Leo: So clearly somebody got ahold of Avast's certificate and was using it somehow. I think because she was working for a Synagogue she was probably targeted. In any event, the reason I bring this up is here's a person who shouldn't be using Macintosh and was probably told rightly, if you're going to use a computer, it's easier, it's safer, use a Macintosh. Even that wasn't secure. Here's the fall. She bought anti-virus software and installed it, and the Mac let her because you're the user, you know what you're doing. She shouldn't have been using a Mac. My position is no one should be using it unless they have the knowledge to do and most people don't.
Greg: Avast attaches to the micropanel, and yeah.
Alex: Leo, I think you make a very functional point, and one that I can't dispute for the short term, but to me, moving people back to the less complex system that may be more effective is anti-progress in technology.
Leo: That's real progress!
Alex: I had it backwards all along. Dang. I still think you're completely wrong. I respect your point and I think you're lovely.
Leo: thank you. Let's take a break on that high note. Thank you, Alex Wilhelm is here from Mattermark.com where he's an analyst for financial markets. What do you do?
Alex: I cover where technology and finance intersect. Acquisitions, IPOs, etcetera.
Leo: Also, I know what Engadget does, I think. Devindra Hardawar, senior editor at Engadget.
Devindra: It's in the name.
Leo: A fine Verizon property. Any changes there because of that?
Devindra: We have a very nice stub at the bottom of our Verizon posts saying that Verizon is neutral and you'll have to take it out of our cold dead hands.
Leo: Rightly so. It may be foolish to do that. They now, or soon will be owning Yahoo. I believe they also have acquired Alta Vista and... AOL and Yahoo they're collecting older properties. Actually it's interesting. Arianna Huffington has decided that she no longer wants to work at the Huffington Post. However, the name will remain behind.
Greg: She's quitting the Huffington Post to start her well-being website.
Alex: Have you heard the naproom story?
Leo: Hold on. We've got to take a break. I want to hear that, and I want to hear that. I want to hear... it's great to have Allyn Malvetano from PC perspective and Greg Ferro from the Packet Pushers Podcast. Packet Pushers.net. Our show today brought to you by My mattress. Speaking of Sleeping properly. Did you know that's what was coming up?
Leo: That was a Segway you were working on?
Greg: No. It was a line.
Leo: Ariana, darling, all you need is a Casper mattress. you'll sleep better you'll love it. I love our Casper mattress, and you know what's great, because I didn't buy it because it was less expensive. It is because they eliminate the middle man. I didn't buy it because of the hundred nights that you would try before you buy. I bought it because it was easy to get inside. It comes in a box. It's awesome! Delivered directly to your door, it's an award winning mattress, just the right sink, just the right bounce. Latex, springy latex foam on top of supportive memory foam to give you just the right amount of cush but firm support. It's exactly what I need and exactly what most people need. The beauty part is there is no risk. You buy it today, you'll get free delivery. Hang-less returns within a hundred day period. You don't have to pay a penny. Just call them, they'll come and get it. This is so much better than lying on a bed in a show room for five minutes, you'll get a hundred days to decide whether it's right for you. I think you will. Free shipping and returns anywhere in the US and Canada. They are made in the US. Business intelligence, innovation award winner for 2016, by the way. It's obsessively engineered mattresses at a fair price. Even fairer when you use our offer code TWiT at Casper.com. Go to casper.com/twit. casper.com/twit. Use the promo code TWiT. You'll get 50 bucks off any size mattress. I got one for my son in college. Dorm mattresses... not such good mattresses. It came in a box, he brought it upstairs, he opened it up. It was pretty cool. It was inexpensive enough that after a few years I said you can leave it in the dorm now. You don't have to take it with you. Casper.com/twit, enter the code Twit for 50 dollars off, terms and conditions apply. Casper, the best mattress, Ariana. Just get it.
Alex: Your Greek accent sounds like Count Dracula.
Leo: I have interviewed her when she was Arianna Stasinopoúlou. When she wrote the book about the Castle, and it was so good, so beautiful. Have you ever talked to her? She talks just like that. Devindra, darling, I have to tell you. I'm going to start a new website called Thrive? What's it called? Huff Thrive?
Greg: Fluffington Post.
Leo: I will fluff your pillow for you. Can I just say that when you get to a certain age it's normal that you don't sleep well? I'm just saying.
Greg: Maybe she could have seen a doctor about it, instead of starting a startup.
Leo: You can't really knock it. She did pretty well with it the last time.
Devindra: It's an important thing she's talking about.
Leo: Less that it's she's not having a good nights' sleep and more because she sees that as a big opportunity. Health and wellness.
Greg: Not to mention verizon is coming and it may look redundant. They're not going to put her on the board or anywhere important.
Leo: It's sad.
Greg: Her ego isn't going to take the battering, and she's going to be out of there.
Alex: When I was at AOL two and a half, you're currently there. Do you ever notice that she worked there. When she announced she was stepping down, I was shocked that she was at all around. That was my surprise. I felt none of her influence internally.
Greg: That's called leadership right there.
Leo: What does John Oliver call Huffington Post? He has a great name for it. This slice and dice and re-use blog. They don't have anything original.
Alex: They do have a decent politics. Buzzfeed news is actually good, but it's wrapped in this cuccoon of other stuff.
Leo: According to Re-Code, Why did Arianna leave the Huffington post? It rhymes with Yahoo, darling. What rhymes with Yahoo?
Greg: It must be pretty sad for her to see her entire life's work reduced to an ad tech. Verizon is making 24 billion a year out of selling data about what traffic crosses their network to add to tech companies. That's today. It's a 24 billion dollar market. By buying the end point, they can take that ad tech data and unify the users. That's why Verizon is out there buying up content properties. Because they've got better ad tech, and they can drive higher users. It's not about the content, it's not about the companies. It's about displaying ads.
Devindra: I thought they were sad about losing out the old web too. Got to get all the toys you couldn't get the first time around.
Leo: That AOL was so fun.
Greg: I can't see anybody on the board of Verizon signing off because we missed out on eleven other...
Leo: That's life.
Greg: They are the third largest ad tech company behind Google and Facebook. They now have 5 to 6 percent of the market.
Leo: They're really not a cellphone company any more. That's not the business anymore. I don't blame them. This is what you're seeing is a lot of companies don't want to be pipes. Dumb pipes is not a long term business strategy. So Comcast bought NBC Universal, they became a content company. Smart Pipes. In fact, you see this horrible conflict of interest in the Olympics. You're not going to experience this in the UK, but here in the United States because NBC owns the Olympics and Comcast owns NBC, it is a terrible Nexus of content plus distribution. We broke up the movie theaters, used to own all the movie companies and we broke that up years ago, because..
Greg: That has to change. For example, I'm here from the UK and in the UK, my cellphone has an unlimited plan. That means 24 hours a day, seven days a week, I can draw data all the way down. What I'm now doing is I can take that phone over here and get unlimited data when I'm here. I don't pay roaming charges. It's just part of the plan. The company just gave...
Alex: What sorcery is this?
Leo: How could that happen? It's not unlimited.
Greg: My guess is that the company in the UK..
Leo: Google has a deal with Three for Google 5.
Greg: Do you know how much it costs to collect the accounting data to do that? It's huge sums. Hundreds of millions of dollars in accounting...
Leo: They don't, there's a mismatch.
Greg: You have to get the data records from... when I was working for a large mobile phone carrier, we used to have to assert these massive pipes for the accounting data to change. We used to interchange. What used to happen was we'd count all of the calls that Verizon were on our network, and then they would count all the minutes that were on ours. When there was a discrepancy, we had to reconcile them and somebody paid each other. You know how much it cost? Tens of millions of dollars.
Leo: That's why data roaming is so expensive.
Greg: The actual cost of it is nothing. Because the Internet traffic doesn't get back to the UK and it just routes out Internet. Verizon has all those Internet exchange points. It's just straight up arbitrash. If I charge you ten dollars to join TWiT and then you charge me ten dollars, paid me ten dollars to appear here, who is winning? That's exactly what they're doing. It's a straight up arbitrage deal. What I suspect is they're saying why am I spending all this money on accounting? Why don't I just call it Squid and say whenever you're in these countries, that's just...
Leo: That was the big announcement here was that Google 5 will charge the same amount in a partnership with three, among other carriers, in 19 other countries. Similar, we don't have unlimited data, it's ten dollars a gigabyte, but that's actually as good a deal as far as I'm concerned.
Greg: It's not like it's a massive problem. Think of how much money you've saved adding the records to your bill. How many customers are ringing up bitching about...
Leo: My iPhone says I used a hundred megabytes, you say I used 400. It happened to me. It doesn't match what your phone says.
Greg: They're charging you seven dollars a meg or five dollars a meg. It's awful. Unhappy customers. Why don't you just grow up?
Leo: That's what's wrong with the medical system in this country. Something like, a huge percentage, I don't remember the exact number so I won't quote it. A huge percentage of medical costs in this country goes to administrative cost. For similar reasons. It's the cost, building the cost of figuring out. The insurance costs, it's just crazy.
Greg: We have the same problem in technology.
Leo: Why don't we just stop charging for everything?
Devindra: I guess it's what T Mobile was working towards with their roaming stuff. And it's fantastic. I've taken that to Barcelona, I've taken it to Taiwan. It's 2G, 3G speed.
Leo: Sometimes it's faster.
Devindra: Sometimes it speeds up. It's enough for you to find a location on the maps.
Leo: T Mobile still tries to get you, at least this is my experience, to buy faster packages. Every day I got a text. It says you want to buy a pass, you can buy a pass for a faster package. It's nice to have some... Sprint has responded in the US with a similar package. Their one world package is similar. That's where competition makes a huge difference. Of course AT&T and Verizon don't have to do anything, because they're wonderful.
Greg: They have a government mandated monopoly.
Greg: Of course they don't denote anything to the political process.
Leo: You guys are so great, I'm going to talk about anything you want. You said there was something... did you say Alex that there was something you wanted to talk about?
Leo: Always the slacker. You must be a millennial.
Alex: I was trying to not talk about the conversation to get better, now you're bringing me back down.
Leo: Just kidding. Actually, you can talk about that T shirt. Turns out no one else was bidding for Jet. Wall Mart bid itself up. How did it get to that?
Alex: Jet sold for a discount.
Leo: Three billion dollars, apparently they're paying to Mark Lord. Lock him in.
Alex: That's the general idea is that he has a very expensive hire, because the company as it currently stands, loses 140 million dollars a month...
Leo: To Amazon.
Alex: No. Just they lose a bunch of money operating...
Leo: Jet does.
Leo: Jet was always a non-starter, I thought.
Alex: I agree. But other people who didn't agree are getting hundreds of millions of dollars.
Leo: Those of you, and that means almost all of you who have ever used Jet... you used it?
Leo: But you have a T shirt. I actually said this before. A couple people emailed me and said, I used Jet once. I tried to use Jet, but it wasn't clear initially. They were going to charge a fee just to buy, then they promised with their ads that they would shrink prices with a special gun.
Alex: Honey I shrunk the brands, I guess. It's a terrible idea.
Leo: No one is going to... you cannot go into Amazon's backyard and beat Amazon. Ali baba could beat Amazon in china. Ali Baba might make a run of Amazon in the US.
Greg: I don't agree with that statement. I think the markets are too radically different. If you've ever seen a Chinese website, it looks like Big bird and the cookie monster.
Leo: I've tried to buy stuff from Ali Baba, and I agree. It looks like a junk store. It's cultural. It's not made for us.
Alex: There was a blog post in an article this week. The first thing saying, there are Chinese companies who are actually building startups for the Western World. Being quite successful in the Western World. Unable to translate them into Chinese properties. There are equally Chinese companies who tried to bust into the global markets and have been spectacularly unsuccessful. I think it was Ali Baba who hired a big foot all starter to do a big push. They spent like 300 million on a marketing push.
Leo: It's hard. It really is hard to cross those cultural... Apple is learning that in China. Their sales have tailed off. Uber pulled out and let Didi have it, right?
Alex: For a nice chunk of the company.
Leo: I'll pull out for a nice chunk. What?
Alex: It's not me. This kid in the middle.
Leo: People say that.
Alex: They don't say it with a twinkle in their eye. There are children here.
Leo: Anyway, we were talking about Amazon and Jet. 3 billion for Mark Lord. He created diapers.com. It was a great story, as soon as Jeff Bezos saw diapers.com coming, he immediately undercut them completely. When their value started to plummet, he said I'll take it. I'm sure Zappos is a similar story. Zappos, unlike diapers.com has been able to run autonomously.
Alex: Yes, but I don't think Jet's remains will be independent inside of Wall Mart, They'll be part of the larger corporate wall mart brand.
Leo: What Wall Mart needed was a digital strategy to survive in a digital era.
Alex: I agree entirely, but why do you have to buy Mark Lore for 3.9 billion...
Leo: That's the question. Is Mark Lore the guy?
Devindra: Have you been to Wall Mart's website?
Alex: I haven't.
Leo: Is it terrible?
Devindra: It's a basic store.
Leo: You know what they need on the Wall Mart website? They need the people of Wall Mart Tumblr. I'm just saying.
Alex: It's a Chromebook on Wall Mart.com!
Leo: For 79 bucks, get the tech for your student. If they mixed the people of Wall Mart, you go to Wall Mart because you want to see the people of Wall Mart. If they mixed the people of Wall Mart into the website... I don't know. People of Wall Mart, they're different. This is the Tumblog peopleofWallMart.com.
Greg: When you look at this, don't you just get the sense that your life is not as interesting as theirs?
Leo: I go to Wall Mart to keep my life interesting.
Alex: That guy has Mad Max/burning man castoff.
Leo: Apparently Ozzie Osbourne shops at Wall Mart.
Alex: If you have a late night by yourself, you can always turn to people of Wall Mart.
Greg: I think partly the Jet acquisition was triggered by numbers. Wall Mart is now smaller than Amazon. Wall Mart the physical store announced they're shutting like a hundred stores. Sears is equally bad.
Leo: Sears exists?
Greg: They're cutting hundreds of stores.
Leo: Sears used to in San Francisco have the big store. It was beautiful. I used to go there all the time. I think it's a bowling alley now. But there is still a Sears somewhere.
Greg: It's part of the research if you read up on Jet.com and why they acquired them is literally Wall Mart needs to react to Amazon, and they've been trying to organically build. They've built up a very strong IT infrastructure team. The value of Jet is its infrastructure can weld. It doesn't lift straight out of Amazon..
Leo: Why is it so hard to build this stuff?
Greg: It's not hard to build it. It's the time.
Leo: It's cheaper and faster to just buy Jet?
Greg: They need to act now and not let Amazon... if they were to build it, it would take 3-5 years.
Leo: You would think Apple, with more than 200 billion dollars in the bank, for instance, could write or create a cloud service that worked, and yet, there must be something hard to do about it.
Greg: Apple's problem is that they're working at such unbelievable scale.
Leo: So is Facebook. So is Google. So is.. not Yahoo. Actually Yahoo is.
Alex: Wall Mart had 4 times the sales of Amazon in the last quarter.
Leo: They aren't small.
Alex: No. Wall Mart is 115 billion, and Amazon was just over 30 billion in revenue.
Leo: Online they're much smaller. They see the end of the brick and mortar.
Greg: If you look at the growth numbers, you'll see Wall Mart is shrinking as Amazon is growing. So the answer is, the problem with Apple is it's just so unbelievable in its scale. If you're looking at what they did what Drive, they didn't do that for five million users or ten million users, they did that for 1.5 billion people.
Leo: I hear three years in it doesn't work very well.
Greg: I'm finding iCloud drive is working... I just switched the bulk of my sink into iCloud drive.
Leo: You are a brave man.
Greg: Steadily away from Dropbox. I've got all three. Dropbox, iCloud Drive, and google Drive. Google Drive is awful.
Alex: Apple can really improve this over time.
Leo: What's bad about Google?
Greg: The Apple crashes after about 2 and a half minutes, so it stops syncing on my Mac. I've deleted it, it takes ages for the Google Drive to sink.
Leo: We use Google drive all the time. Do we have problems with Google drive? What are you doing with Google? Oh. You're syncing files. I understand what you're saying.
Greg: We sync our podcasts. We drop the files and send them off and Google drive is... they're not doing it right either.
Leo: It's hard to do.
Greg: If you're going to point the finger at Apple, you've got to think Google. Amazon has got a volume drive syncing engine. Nobody uses that for the same reason.
Devindra: You guys were asking why is it so hard for Apple? It almost seems like they constantly under estimate the work involved. It's not just the scale and the resources, It's like let's just get into maps. That was a disaster and we heard this week what the Maps fiasco led to the public testing for...
Leo: NEQ said that in the interview. Singly devoid of any information except for that.
Alex: But that was a moment of real honesty. I appreciate that. As an Apple observer, I have no insight into how they operate because they are notoriously private, but that was them saying here's a problem we have, here's how we fixed it. Which is a big ticket for Apple.
Leo: They said as a result of the failure of Apple maps, the public debate is going forward. They have a better job of the updates of iOS 10. Arguably in better shape than any version of IOS prior.
Alex: How long was Gmail in Beta? Do you recall?
Leo: It's still in Beta.
Alex: That joke was funny for a while.
Leo: It was almost ten years. It was a long time. That's just a badge.
Alex: It's always just a badge.
Leo: What does that mean?
Alex: Alpha and Beta are value decisions by people who make the product.
Leo: After a year, it really says we just don't have that much confidence in the product.
Greg: I've built infrastructures like what Apple is doing with iCloud. It's not easy. Apple made some mistakes in the early days. They bought off the shelf from existing vendors, they bought terra mark for storage.
Leo: They were using Azure as far as I know.
Greg: They use all different cloud services. A lot of the iCloud photos don't have Google at the moment. Some stuff comes off Azure, and a lot of stuff comes off Amazon in addition to what they're using. The challenge is that they can't scale up fast enough. They can't build data centers fast enough. They've also got a problem with...
Leo: Maybe they should stop trying to push people into using iCloud so far.
Greg: The challenge here is if you're going to put a billion users into something, you need a couple data centers. They take ten years to go through buying the land, getting the approvals, getting...
Leo: That does make sense. When you're talking about infrastructure like that, it takes a lot.
Greg: What they did was to keep moving at speed. Apple was moving quite quickly a few years ago, they were out sourcing their tech. They were buying it from other people and letting their suppliers, integrators come and provide those services. Those companies didn't do a very good job, because they couldn't do it at Scale. Apple's turned around and started developing all of that infrastructure to scale in the way that Google, Amazon, and Facebook.
Leo: Who does it best? Is it Amazon?
Greg: Probably Facebook.
Leo: Facebook does seem to be pretty reliable. But you don't sinc gigabytes of data to your Facebook cloud.
Greg: Facebook does for you. Facebook is unique in its ability to change its platform. Amazon does a great job, but that website is a usability disaster. It's like a dog vomited straight onto your web page. If you want to bark at a pair of glasses, how do you go shopping for a pair of glasses, unless you know exactly what the product is? It's like going diving through a dumpster.
Leo: Like Ali Baba.
Allyn: Even the search is pretty bad.
Greg: Going shopping on Amazon is great as long as you know what it is. Otherwise it's like shopping at Cosco. It's somewhere in there. The similarities between Amazon and Cosco aren't exactly unknown to me, but they're the same business model.
Devindra: I almost wonder if they're spending more time on their mobile apps than their website at this point when it comes to usability. It's so much easier to pull up the app and get something and finish an order rather than...
Leo: It makes sense. That's how they want people to shop and people want to shop.
Devindra: That may be part of Wall Mart's decision to purchase Jet. Wall Mart, anybody who needs to go to the store, they say, "I'm going to go to Wall Mart." It's a place you go to get stuff. Now you're buying stuff online, you go to Amazon. That's the only place most people go unless you want specialized hardware. If Amazon wants to compete at that level, yeah. They need better infrastructure.
Leo: This is a classic challenge is taking a brick and mortar business and moving it to the cloud. Very few have done this successfully.
Devindra: Jet's thing too was about discounting items. Wall Mart has the idea of being cheap. If they can combine those brand features, that could be interesting.
Leo: Do you think they have a chance?
Alex: That's my question. What is the chance Wall Mart pulls of a reasonable Amazon. Not number one, but actually a viable long-term brand that people use after the Jet acquisition? I don't think it turns out that way, not to be negative.
Leo: But how loyal are people to Amazon? If you suddenly found a good experience that offered a good price, wouldn't you be willing to shift? I don't think I'd have a problem buying from another store.
Alex: Right, but I don't think they'll pull off that level of service.
Leo: They have to do that, of course.
Alex: I don't really like Amazon all that much. I use it because it works. There's no other Amazon equivalent in the US.
Leo: Also, we're bought into it. We gave them a hundred bucks so we'd have Amazon Prime. We've committed to it, and I think there's a perception that Amazon has the lowest prices, which I think is in fact not true.
Devindra: It's more about guarantee of service. If I know I need this cable tomorrow, or even today, Amazon has enough options, I know I can trust Amazon to get me a...
Leo: That's actually the big selling point. That's why those fulfillment centers are so important. The infrastructure, the logistics. You might say that Amazon is a logistics business.
Alex: So is Wall Mart in a lot of ways.
Leo: They have RFID in all their products, they're very smart about it.
Alex: Wall Mart squeezes money out of their supply chain very efficiently, which is how they have profit margins. More power to them.
Greg: They've moved away from holding stock and they're more about selling the goods. Amazon doesn't want to keep building its stores. What it wants to do is be a platform, so that when you go to shop, pay your hundred dollars a year, and that's what they want to make the money. They don't actually want to make a profit on selling the goods as much. They want the hundred dollars a year out of your pocket. That's why they give away movies and Kindle books basically as part of the package. If you could get a billion people paying you a hundred dollars a year to use my platform, what have I just done?
Alex: Created a hundred billion dollar revenue stream.
Greg: Yeah. If I don't have run the warehouses and don't have to pay for the stock in the warehouses, the people who manufacture the goods put the stock in my warehouses, or maybe ship it directly, but using my distribution system, then all of a sudden, my mechanics are I'm making a hundred billion dollars for what? Running a website.
Leo: In five years is it going to be Wall Mart and Amazon? Is there anybody else? Is Target in the mix? So really...
Alex: No. Target got hacked and their stores don't work.
Leo: Really it's going to be Amazon and maybe Wall Mart.
Alex: Target's hack doesn't matter.
Leo: That doesn't stop people from using it.
Alex: Not at all. Their share price is up 80% since...
Devindra: The only reason you go to Target is to walk into a slightly nicer Wall Mart is all.
Leo: It's a slightly nicer Wall Mart, really? We don't have a Wall Mart around here. That's why I don't...
Devindra: The lighting is better.
Leo: Lighting is better?
Devindra: Lighting is better.
Devindra: Same cheap prices but there's a little more style.
Alex: The chatroom is loving this by the way.
Leo: Devindra is right on.
Alex: God, Greg is good. Keep coming back, man.
Leo: And who? Wilhelm who? Alex who?
Alex: I know they don't like me, so.
Leo: All right. Let's take a break. I do want to talk about a lot of other things. You guys, I think you're doing such a good job you should pick the stories you want to come back with. So that's your assignment while I, while I, the rest of us we'll talk about Audible.com. And if you want to talk about Audible you can too. It is very distracting whenever I talk about Audible. We all end up talking about our favorite audio books. I spend so much time listening to books on Audible.com. I've been an Audible subscriber long before they became an advertiser. Although they have been the longest advertiser I think, one of the longest advertisers on TWiT, for almost 10 years. I think 8 years, I think something like that. I've been an Audible subscriber for 16 years, hundreds of books. I love Audible and I think you will too. It is the world's largest audio bookstore, a quarter of a million titles and not just novels, but fiction, non-fiction, college courses, everything you would want to listen to. Audible.com. If you go to Audible.com/twit2 we'll set you up with the platinum plan. This is free for the first month. They'll give you 2 books, or 2 credits. Some books are, some of the really long books, one tenth of one percent are long enough so they are 2 credits. But you'll get to use those 2 credits towards any book or books you want. You'll also get the daily digest of the Wall Street Journal or the New York Times, yours to pick, every single day. And then after a month you have to make the decision, "Do I keep going or not?" But if you decide to cancel you'd have paid nothing. Those 2 books will be yours to keep forever. It's a great way to try, something I know, some of you are like "Hmm? Do I want to listen to audio books? What would it be like?" Trust me, I love them. I am as you can tell, an Audible fanatic. Listen all the time. I'm getting my son to listen to. We were talking to Henry earlier, my college kid. I said—he said, "I want to learn." I said, "Well you got to read." Then he said, "Well I get sleepy when I read." I said, "Listen to Audible." He really loves audiobooks now. He's listening to a lot of audiobooks. Go to Audible.com/twit2 and pick a book or two. You'll get them for free for the 1st 30 days. Enjoy it and let me know what you think. I think you're going to become an Audible fanatic. Audible.com/twit2. If you are a Times subscriber you probably got The Underground Railroad last week it came in the Sunday times, a fairly long excerpt. It was kind of a neat deal. It's maybe because it's the return of the Oprah Book Club. It's the first book in some time. Brand new book by Colson Whitehead, that's also on Audible. So if you read the New York Times excerpt and you want to read more, you could buy the book but you'll get every word brought to life. That's the best thing about audiobooks, at Audible.com. And this could be yours for free, The Underground Railroad. Audible.com/twit2. Try it today.
Leo: We're talking about the week's tech news. Allyn Malventano's here. You were here for the—what is it? FMS?
Allyn: Flash Memory Summit.
Leo: So what is that?
Allyn: So it used to be, 2, 3 years ago and before that, anything that had to do with solid state drive development, specifically announced at the Consumer Electronics Show in January in Vegas. And over the past few years, that whole storage subsection, like Flash related or anything solid state storage related has kind of shifted to what is the Flash Memory Summit which is actually down here at the convention center.
Leo: Ah. Because this was—IBM's facility in South Bay. Wasn't that where they were doing their SSD, their very famous SSD?
Allyn: Probably developing the way early stuff.
Leo: Yea. And then Intel of course was the leader in SSD as well. So it makes sense that Silicon Valley would be where you would want to go.
Allyn: That's where a lot of the development is, a lot of the—
Leo: Although isn't Samsung now kind of the leader in SSDs at this point?
Allyn: And they fly right over here.
Leo: They just come over (laughing).
Allyn: And they do keynotes. Samsung frequently has their own, what they just call a SSD Summit and they do that in Seoul. And I've probably been out there 4 or 5 times by now to cover announcements from them. But they also do announcements at the Flash Memory Summit here.
Leo: It's funny because in the early days of Flash, you have the whole chart. There were all sorts of questions.
Allyn: The SSD Decoder.
Leo: The SSD Decoder. It was very complicated to figure out what firmware.
Allyn: You thought there were too many SSDs to—
Leo: But also what I think what's happened is they've become commoditized now and now the main brands are all pretty good I would guess.
Leo: No? There are really, there are significant differences?
Allyn: There are. And I ended up having to develop new testing just to show those differences better.
Leo: What are they? Speed, reliability, what are the—
Allyn: You'll end up running into—most of it boils down to—the Flash itself is typically fine. It's just how the manufacturers implement their firmware.
Leo: Right. So it's still that. That was the big issue for a long time.
Allyn: It's all about managing how the Flash is written, too, because it's kind of like a musical chair problem.
Leo: Things like wear leveling.
Allyn: Yea you have to do wear leveling. You can't write very small, random to it continuously because it likes being written in large or erased in large chunks, relatively large chunks. Like Flash for me, the analogy I usually use is like a bunch of little CD-RWs. Remember a CD-RW? If you filled it and you wanted to erase a file, or overwrite a file, you had to basically copy the whole thing off, wipe the whole disc and start over. Flash has to do that in like 6MB chucks. So you can't just do very small, random writes to it continuously. So there's a lot of behind the scenes jockeying and juggling around of data that happens that you are completely unaware of, it just all happens. Like there is an S at the end of that system. It's doing that right now.
Greg Ferro: That's why we have SSDs with 8GB of DRAM for removal.
Leo: Just so they can swap these 8 gigs out.
Greg: So the DRAM actually works as a multi-tiered cache inside the SSD, but also gives them read write space so that you can actually juggle it and hold it before you write it down. And that DRAM is actually battery backed so if the drive depowers, your memory is safe.
Leo: Also part of the problem, the fact that operating systems don't know about, they still think in terms of cylinders and sectors and all of this stuff.
Allyn: It's all obfuscated now. It's just a sector number that you write to and that's—
Leo: So the OS is no longer the problem.
Allyn: That's not really the problem especially with NVME and newer technologies that are just designed for—
Greg: NVME is very exciting because it does away with all the sort of scuzzy sectors. It starts to say, "I want to pull off the data."
Leo: It's an address.
Greg: You remember the old, scuzzy protocol which is basically just fetch this block and you serially fetch block after block. NVME says "I need this data block," and the drive just— you remove all of the scuzzy-
Leo: But some of the intelligence is now moved into the drive, right?
Greg: It's in both. NVME simplified the protocol between the operating system and the drive fetch. And then the drive itself designs how it wants to do its housekeeping. But what, Allyn, we should probably talk about 3d NAND and what's going to happen when you get the real innovation.
Leo: Did they make some announcements at FMS about that?
Allyn: Samsung did.
Allyn: So Samsung has been iterating on their 3d NAND since it came out a few years ago. And there was actually a 24-layer V-NAND, it was version 1, but nobody saw it in any consumer parts. It was kind of just like trial runs.
Leo: You're going to have to explain this to me. Is this like RAM or stacked RAM and 24 layers of it? What does that mean?
Allyn: So there's different ways of stacking things. There was actually somebody tweeted at me earlier today, and they were kind of confused on like, "Ok, what does this mean? Does this mean I'm putting 64 dies stacked on top of each other?"
Leo: Yea, what does that mean?
Allyn: No, this is actually—
Leo: Within a single die.
Allyn: Within a single die, 64-Layer announced at the show, 64 layers worth of cells that are each storing charge to store data.
Leo: Does that mean it's denser?
Allyn: Because if you're able to go vertical, that reduces some of the limitations of what used to be the issue. Actually if you scroll down a little bit—
Leo: How do you address this? Is it addressed in XYZ?
Allyn: It's still the similar kind of addressing. It's sort of like an X and Y, like a row and a column kind of thing.
Leo: But now you've got depth so what do you?
Allyn: It's obvious. It's a big row. You don't address like a specific height or whatnot.
Leo: (Laughing) I see. All right. Ok.
Allyn: But there's actually a cutaway, if you scroll down a little more on the page, of what that stuff looks like. I think. Or no, they already passed it. It was further up. Anyway.
Leo: So you can actually see into the die. Is that it?
Allyn: That's a cross section. So that's a cross section of I think 48 layer, when that came out, the guys at Tech Inside always like cut dies in half and do electron microscope stuff and look at it. So what you're looking at is rows worth of cells. And so in this case, Samsung frequently runs that in TLC mode. So each one of those little squares, right, is holding 3 bits of data. And then you multiply that out and you've—
Leo: So what this means is we're going to get more capacity in the same space.
Allyn: Yes. And along those, that same exact vein, last year at Flash Memory Summit, we saw a 16GB, 2.5" form factor SSD for Enterprise from Samsung. And this year they showed that same exact product.
Leo: Sorry, just ignore that. I don't know why my phone is talking to me.
Allyn: So we saw that same exact product. It almost looked like nothing had changed at all. However, they had now the double capacity per die of this new 64-layer flash inside that drive, so they were able to double the capacity of that SSD card.
Leo: So how big is that, a 3.5" SSD?
Leo: 2.5" How much is that going to hold?
Allyn: In Samsung's case, 32TB now.
Allyn: Sorry, if I said GB.
Allyn: It was 16TB last year, 32TB this year. Big number for a 2.5"
Leo: And the same size as my 950 EVO now?
Allyn: No, no, no, no. Oh, well 850 EVO. But it's a little thicker.
Leo: A little thicker. You had me until you said that.
Greg: I think the most interesting storage that's coming out is the new Intel Optane.
Allyn: Yes, so well, Intel's branding is Optane. Micron announced their branding as QuantX at the show.
Greg: The key thing about this memory is that it's permanent. So unlike Flash, which is you write it and then—
Allyn: That's permanent too.
Greg: Well it is, but—
Leo: But not forever permanent.
Greg: Speed is 100 times faster than SSDs. So now it actually replaces DRAM and you can start to read and write directly from your hard drive.
Allyn: This is called XP oint.
Greg: 3D XPoint.
Leo: Well aren't I reading and writing directly to a hard drive right now?
Alex: You're looking at the wrong side of the table, so.
Leo: I want to be on this side of the table (laughing).
Greg: Think about it this way. Today, you have an iPhone. It has a CPU. Your smartphone has a CPU. It has DRAM and it has SSD.
Leo: Yea, I understand DRAM, memory and storage.
Greg: What if I could replace my DRAM with a permanent storage.
Leo: So it's as fast as DRAM?
Greg: It's about 10 times slower than DRAM. 100 times slower than DRAM but 1000 times faster than a SSD.
Allyn: So that's where—
Leo: How's it going to be treated, as DRAM or as just really fast storage?
Greg: Well this is what we can't know. This is where micro panels come in.
Allyn: That's the—we're on the ground floor as far as it's a new type of memory. If this stuff came out 15, 20 years ago, that laptop would be architected differently than it was right now.
Leo: But I mean one way you could sell it is you're going to have 10 times faster hard drives.
Allyn: And actually that's what we saw at the show. Micron showed a prototype and they actually showed some results, some actual test results of the—
Leo: But the problem is we're connecting our hard drives by buses that are saturated as it is.
Greg: Think about it with IOT. They're talking 50 billion devices in IOT by 2020.
Leo: And they don't have to be as fast as DRAM. They can use—so you can have—
Greg: So why would you have DRAM and an SSD? Why not just have something that is permanent storage but closer to the speed of DRAM and then it clocks out. Now think about the savings in battery power if you've got a 5000GB NVDM 3D XPoint. Then you don't have a hard drive, you don't have DRAM, you've just got a CPU.
Leo: So it sounds like the PC architecture is about to change dramatically.
Allyn: To take advantage of this it would have to. So it's out now, or Intel plans to—
Leo: So basically everything you've just bought is obsolete.
Allyn: Well, it's going to hit Enterprise first.
Leo: Oh, we didn't mention. This is going to be horrifically expensive probably, right?
Greg: It's reassuringly expensive. If it's not expensive, it's not worth having.
Allyn: It's going to sit between Flash which is in your SSD and DRAM. Its's going to sit between those two and performs between those two.
Leo: You know it's interesting. We always think of Moore's Law as applying to computation. But there is also, it's not exponential probably, but there's a very fast growth rate in storage capabilities, in capacity and—
Greg: It's not just the storage capacity. That's true. There's also if you start to take away, DRAM, like temp DRAM at high speed and low speed permanent storage, SSDs and replace it with a single module, then you've actually simplified the operating system. So your power consumption drops.
Leo: You don't save files anymore.
Greg: And then we start to use NVME which is a way more efficient access protocols. So instead of fetching the data from the SSDs as blocks, which means we have to emulate the blocks in software in the drive, so now there is a CPU and 8GBs of RAM. If I could—
Leo: So a new computer.
Greg: All of a sudden I've got—so I was wanting to talk to you a little bit about 5G and the upcoming changes to that. So we talked about 4G. One of the—
Leo: Don't get us on the 5G now. That's going to be a whole other thing. We'll get there, but ok.
Greg: What we're heading towards is devices that run for 10 years on a single battery charge that are connected to the 5G network.
Devindra: That's interesting. We were talking about the Google stuff too, like Fuchsia as well.
Leo: Now it all kind of makes sense, right?
Greg: There's a convergence play going on. Operating systems are getting smaller. The technology inside of our computers, the storage is changing which means we would need new operating systems that don't rely on reading and writing to DRAM and then flushing to disc.
Leo: So what is that world look like now, where we have—I mean it really does facilitate the idea o distribution of computing power to small devices all over, yea? Intelligent devices everywhere. You've got machine learning now and capabilities to make these devices more autonomous, more intelligent, able to kind of operate—in fact this really facilitates the autonomous device in some ways.
Greg: The way I look at it, I'm a networking professional so my mine job is how does the network connect things together. And I see what today we have most of our computers as personal things. Your computer talks to my computer. I talk to my server in the cloud. I talk to Facebook. That's person to person. Or person to machine. IOT is much more focused on business to machine or machine to machine. So today you have no sensors in this building watching temperature except for cooling.
Leo: Right (laughing). We have thermostats everywhere. John is very defensive about his studio. Don't attack his studio.
Greg: But imagine if you had a sensor on every door so you could track every time it opened. And if you had sensors at every point throughout the building checking the temperature and also light sensors and then also motion sensors and you had a software platform that could bring all that together and say, "There's been no motion in this room. I'll turn the cooling down and turn the lights off. And I'll also lock the doors because there's nobody in the building."
Devindra: You'll open the pod bay doors too.
Greg: Yea, etcetera, etcetera. And all of a sudden, today we don't do that—
Leo: Or not.
Devindra: Or not.
Leo: Or not.
Greg: And today we don't do that because linking the networks up between the cooling system and the lighting system—
Leo: Actually it's the weirdest thing in the world because my Tesla has some of this. When I walk away, you don't turn your Tesla off or lock it. You just get out of the car and walk away. And it knows what you've done and it locks itself.
Greg: And it communicates back to Tesla, it uploads to—
Leo: By the way, Tesla knows what it's done as well.
Greg: And there's a new form of Ethernet inside cars. So they're using, there's a 1GB—
Leo: They're finally replacing CAN bus with something.
Greg: There's a new version of Ethernet called Car Ethernet designed to run in electrical, noisy environments, etcetera, etcetera. It's all been ratified and tested out and the new BUS is now just Ethernet. So you're now running Intel CPUs.
Leo: I didn't realize how noisy the electric motors were. I can't have an AM radio in my Tesla because there's so much noise from the electric motors. So I hadn't even thought about that but that's got to really impact the network as well.
Greg: Yea. So they announced moving to a standard. They're moving away from 10GB Ethernet to 1GB Ethernet so you can now—and then putting sensors in the taillights. So if you taillight goes out, that's actually a sensor, not in the bulb, but a separate light sensor to say your taillight is no longer lighting up.
Leo: Soon it will be the bulb.
Greg: And soon it will be—well, maybe it will be the bulb. I'm not sure. But however you do it, we've got this whole sensor revolution. What if I could start putting sensors out there that are just connecting to the 5G network and there's new 5G standards like NBIOT that are designed to be this very low power, literally uses CR20T3 like you know, those little tablet sized batteries, those little flat things and run for a decade, on a sensor for a decade. And that's that machine to machine. They're talking 50 billion devices. Well maybe 10 billion of those are battery operated, or 20 or 30 billion. All pushing data out into the cloud. So this is where these new memory architectures combine with the new ARM CPUs, these new low power CPUs to—you know Raspberry Pi is the thin edge of that wedge. What if you could turn those into something smaller like a dime or a coin sized computer that has a battery in it and then you know, has a sensor built into that so we're looking at things like silicon photonics where you can build a laser straight into the CPU die and now all of a sudden you've got a laser sensor built into something the size of a coin you can just stick to the door.
Leo: Last week we had somebody on The New Screen Savers who has a rice sized, a rice grain sized computer with memory, CPU, storage and battery, the size of a kernel of rice. You can eat it. It wouldn't be very tasty, but you could eat it.
Greg: Do you have to, is it—do you retrieve it afterwards, like is it expensive?
Leo: Well, right now they're making them kind of—probably expensive and you would want to retrieve it. The idea is these could probably very easily made. What university was that? I'm trying to remember what we were talking about. It was very interesting.
Devindra: So like all the privacy discussions we're having now around IOT—
Leo: It's moot.
Devindra: Well, we're laying the groundwork because in 10 years, it's going to be insane. It's going to be madness.
Leo: Sensors everywhere.
Leo: That's why we have IPV6 as well because we're going to need all those addresses.
Devindra: Well it's kind of interesting too because I remember reading like Kurzweil back in the 90s and Michio Kaku and those folks, the futurists, who always expected that everything would be microchipped by 2020-something and it's funny seeing the groundwork being laid to get us there, and terrifying.
Leo: I would like to talk more about 5G, what it means, the future of 5G, how soon we're going to have 5G, that's something I think a lot of people—
Devindra: That's like a whole show, honestly, the race to 5G.
Leo: I know. I think it is. Yea. The race to 5G and what even 5G means. It was the University of Michigan, and we talked to David Blaauw and it was—this was the computer you see on a quarter.
Leo: It's the size of—it's smaller than the word liberty on a quarter. It's pretty remarkable.
Greg: Well I think—so one of the things that we had listed to talk about here is why Apple's MacBooks are so behind the curve.
Leo: Well let's talk about that in a bit. That's something we can all talk about.
Alex: Unlike the last 10 minutes of which this side of the table was silent. Because there's nothing I could have added to that.
Leo: No, but you understood it and I think it's fantastic, it's fascinating isn't it? I mean—
Alex: Yea, no.
Leo: It's simple. Better, faster, cheaper.
Devindra: Combine storage and RAM. That is fascinating.
Leo: Yea. Yea. And IOT clearly is a revolution that is about to happen even though we really don't have the infrastructure to handle that.
Devindra: It's funny because every IOT project or product I've tried at this point has completely failed on me. I don't know why.
Leo: They're ultimately disappointing.
Devindra: I don't know why. I'm using a Weemo Air Filter right now in my bedroom. Set a time schedule in the app, come in the next day, it's running at not the right time. At a completely different time like at the epic of the time stream. So like something is just not even working there.
Leo: It is an interesting question. What's going to have to happen between this vision of the future and this current situation where—and the things that do work, like my Ring doorbell works, my Hue lights work, but they don't interoperate well at all. The glue that holds them together is wonky stuff like If This Than That. We are in clearly a very primitive stage of all of this.
Devindra: For sure.
Leo: And as usual, hardware is outpacing software. The capabilities of the hardware are rapidly—
Greg: What you're seeing I think is a convergence of multiple advances and you can start to plot a pathway forward. And then you can, people like Alex can then start to turn that into revenue predictions. How do we see that and where does the disruption occur? So if you look at a smartphone today, it's screen, it's CPU, DRAM, SSDs, what does that look like if suddenly you take the DRAM and the SSD out and replace it with a 3D XPoint permanent memory storage? Does that make it more efficient? It might get a bit slower, but the power consumption will—
Leo: It also continues the trend really to move away from discreet CPU based processing units to kind of a wider mesh intelligence that—and sensor driven mesh intelligence. And in that case speed of any individual component may not be critical. The speed of the network will be however. And that's where I'm intrigued.
Greg: Well 5G, 5G is not about speed. It's about much, much more than that.
Leo: Oh, all right.
Devindra: It's just pervasiveness.
Leo: It's more than speed. We'll talk about it in a bit. We're going to take a break right now. Greg Ferro, it's great to have you. It's wonderful to have you in the studio from the Packet Pushers Network, packetpushers.net. He's the host of—is the show called The Packet Pushers?
Greg: Yea, we have 4 channels, doing different technologies. We talk about data networking.
Leo: Don't say 4 channels. That's not good.
Greg: (laughing) But mainly we do data networking and a bit of a new show which as you might have guess I'm a little bit of a nerd. We deep dive on the technology side.
Leo: Sounds good. Sounds good.
Greg: And we also have a cloud based show for those who are into that called Datanauts.
Leo: Packetpushers.net. Allyn Malventano's also here. Allyn has a checkered past but we won't hold it against him. Working for the NSA and a submariner and you were a nuclear officer on the submarine?
Allyn: No, not an officer.
Leo: You just handled the plutonium.
Allyn: Yes, I just carried it around. It was in my pocket.
Leo: (Laughing) Carried it around.
Alex: It was in a special bag. Come on.
Leo: All I can say is he knows how to pronounce amarisium which is more than I can say for the rest of the panel.
Alex: Ok, ok, ok.
Leo: He's also a host in This Week in Computer Hardware which is a great show on this network, TWiT. And writes regularly for PC perspective. It's great to have you on.
Leo: From Engadget, senior editor, Devindra Hardawar. Always a pleasure to have you on the show. Have you got the Note 7 yet?
Devindra: I've seen it. So we have it in the office. Chris Velazco was playing with it. It's a nice device.
Leo: Yea, I want to talk about—I just got the shipping notice so I'll probably get it in the next few days.
Devindra: It's very nice although I think the idea of the Note is just kind of over.
Leo: Oh, we'll talk about that and also about this very nice laptop you're using. First laptop I've seen with an OLED screen.
Alex: It's really nice.
Leo: Really intriguing. And Alex Wilhelm which is just some guy.
Alex: That is correct.
Leo: We don't know who he is. No, he is of course a senior writer at Mattermark. He's editor-in-chief. Forget senior writer, it's the editor-in-chief, my friends. He's the dude at Mattermark.com.
Alex: I should get business cards with The Dude on them. That will last about a week.
Leo: The Dude who bides. Do they have a sense of humor over there, you know if you did that?
Alex: Yea, that would be fine.
Leo: Good. I'm chief TWiT.
Alex: Oh, so many jokes that are not appropriate for this show. It's just—it's amazing consistently how the show sets you up for a joke you can't make.
Leo: (Laughing) yea. Now you feel my pain.
Alex: Oh, I'm living inside your bucket.
Leo: Thank you. Thank you, thank you. Our show today brought to you by Squarespace. Squarespace makes it easy to create a beautiful, beautiful website that runs perfectly because they are the hosting service and they run the software. That means they always keep it up to date. They always keep it responsive. Whether you need a simple cover page, and they have a great cover page facility. That's how I got started with Squarespace, or you're a photographer who needs a portfolio. Or you're a blogger or even a publication, or an online store, it's all built in to Squarespace. You start with their amazing templates, designed not only by the best designers, but also by the best engineers. So the template is bringing to your site state of the art engineering like web responsive design and e-commerce all done beautifully within this aesthetic that you can completely control and make your own. Every site looks great. And even if you have no skills at all, you can do it. Just drag and drop, point and click. But the nice thing about Squarespace is it goes deep. So if you really want to get deep into CSS, you can. They even have blank templates you can design on. There's actually a very large group of developers who work with Squarespace. That's what we used for the artisanal agency site. Lisa wanted to have a gorgeous site. We went to Squarespace and they have a directory of all the people that can do this stuff and for a very affordable price, we were able to get a gorgeous site designed. It works exactly as she wants. So do it yourself or get somebody to do it for you. You're going to get the website you deserve. They have new magazine style templates that would be great for starting a publication, includes all the things publications make a big deal about. Grid style landing pages, infinite scroll, related posts at the bottom, you get author profile pages. There's even an integrated search field in the header so your readers can get right to the content they want right away. And of course, as with all the websites, all the designs on Squarespace, mobile comes with it. So you don't have a separate mobile site. Your site looks great on any size screen. You get a free custom domain if you subscribe for a year. And they now have domain registrar capabilities, over 200 TLDs. Not just .com and .net but all of the fun TLDs so you can buy a domain, an @ Squarespace domains, register it and they'll give you, unlike anybody else, a free, a spam free parking page. So you get a very clean site ready for you until you're ready to start building. And you've got to try the cover page. That's how I got started and I said, "Well I should do a blog. I should do a portfolio." And by that time, I really had a great setup. I'm just thrilled. Squarespace commerce, every template has commerce built in. They recently introduced advanced analytics for commerce allowing you to see the traffic of visitors in real time. That's so cool to watch as people come and go. You can see who abandons the check-out carts and where they do so you can kind of make it work better for your customers. Device filtering and a whole lot more. Anybody who's built or contributed to 3 or more active Squarespace websites now had access to something brand new, the Squarespace Circle. Membership gives you advanced guides, optimized support, 6 month trial periods for new products, projects I should say, and a lot more. Squarespace is amazing. Join the million Squarespace customers who are using it today. Start your free trial, Squarespace.com and the one little thing we ask if you decide to buy, use the offer code TWiT, you'll get 10% off. Squarespace.com. Make a beautiful website today on Squarespace.
Leo: We're talking about the week's tech news. 5G race. You said the 5G word and I'm very intrigued and I think we all are, right? We've heard this. Although we've been through this before, remember how disappointing WiMAX was. 4G. What was that? What did it do? Did it improve things? I don't know. LTE stands for Long Term Evolution. Is that a technology or a marketing term? So I always worry when I hear 5G that they're trying to sell us something.
Devindra: It's pretty crazy. I mean I was at Mobile World Congress this year and this is I think the first year where we started to see the 5G market kind of start to take shape. Intel's talking about it. QUALCOMM was there. Everybody like had big plans. And the fascinating thing too was that I brought up WiMAX to the Intel people, right? Because that's their big mistake.
Leo: One of their flops, yea.
Devindra: They just like they—
Leo: Almost put Sprint out of business, frankly.
Devindra: Totally. So and all the Intel people were like, "Yea, we're not going to do WiMAX again."
Devindra: We're working with the standards.
Leo: But I remember, I mean remember, I think it was on The Screen Savers, talking about oh WiMAX is going to change everything. It's going to bring fast internet to the rural areas. It's really going to make a difference. I don't know what went wrong, if it was a technical issue, a marketing issue.
Devindra: They didn't work within. I think that was the big thing.
Leo: So it wasn't really—
Greg: It didn't achieve critical mass.
Leo: So there wasn't anything wrong with the technology. See, that's why—Intel doesn't really get to say, "We're not going to do that again."
Devindra: Well, I mean, I think what they're saying is "We learn from our mistakes. We're not going to be stupid and we're going to work with partners and other companies and we're not going to try to force our proprietary solution when the market is moving one way. We're going to try to move with ti."
Leo: So there isn't an Intel 5G.
Devindra: Yep. No.
Greg: No, no. And they also learned from Qualcomm. Do you remember Qualcomm 2.5G? It was only in America. When you left America your phones didn't work. Do you remember that?
Leo: I don't. I mean I've seen 2.5G. Was that ASPA or what was that?
Greg: I think when AT&T had phones they were using 3G if you had an AT&T phone you could walk off to Europe and it would work. But if you had a Verizon phone. What happened was that Verizon was using Qualcomm technology and the challenge was that Qualcomm took a fairly high percentage in profit margins in terms of licensing fees associated with the technology. And 3G was open and no one had to pay patent fees. And of course if you are making technology equipment, guess which one you went for? You didn't go for Qualcomm's equipment.
Leo: So what's driving 5G besides the desire not to screw it up?
Greg: Well basically there's several different trends. So 5G isn't one thing. It's not just speed, although there are some 5G standards that are focused on delivering 1GB performance.
Leo: So these are—so there is a standard body and will have a standard. Even the definition of LTE was fuzzy.
Greg: Yes. The original idea behind Long Term Evolution was it was going to give us a foundation of where the standards could iterate and again though, it turns out that voice doesn't matter, data does. And you could—
Leo: Everything's data now.
Greg: Everything's data. And there's no point in having—
Leo: Including voice.
Greg: Yes, that's right. So running voice over IP doesn't—makes more sense than trying to run voice in a dedicated frequency.
Leo: By the way, is it pronounced VoLTE or VoLTE?
Greg: Who cares?
Greg: I don't know. I think people who have implemented have sort of regretted the decision, right? Because voice doesn't—
Leo: Voice over LTE.
Greg: People don't use telephones to make calls.
Leo: They use data.
Greg: They use data.
Devindra: But you know when I do get a VoLTE call, it sounds amazing.
Leo: It does sound good, doesn't it?
Alex: It's almost creepily high fidelity when you use it.
Leo: Yea, like are you in my head? How did you do that? You're not used to a phone sounding normal.
Alex: Right. I want to hear you breathe into my ear, and yet that's what you get. Well, Leo, you can breathe in my ear.
Greg: You're using telephones for phone calls. I mean, what the hell? What sort of freaks are you that you use phones for calls?
Devindra: I've got to call my parents.
Alex: I keep the phone on my iPhone off my bottom deck. And I feel like it was a big moment for me. Like I don't use this at all. So I just put it in the middle of the rest of the apps.
Alex: Yea, yea, yea, I'll show you.
Leo: Wait a minute. You don't have a phone call icon on your dock?
Alex: No, I turned it off.
Leo: How do you know what to do?
Alex: I just don't do it.
Greg: Well I don't use the phone at all. Everything I do is on an IP conference.
Leo: So this panel is living in the future. Oh, you aren't living in the present.
Alex: This is the future? The future stinks everyone. Disappointments are incoming.
Allyn: Go back, go back.
Leo: The rest of the world still makes phone calls right?
Alex: I mean I call like my parents and my family. But everyone else is like Twitter or whatever.
Greg: My problem is I live in the UK. It's an 8-hour time difference to the US.
Leo: Right, so you don't call anyone.
Greg: So I get interns ringing me at 2:00 in the morning, right? So my level of hostility towards telephones is fairly high.
Leo: But it's not like email's any better so what do you use? Texting?
Greg: I use a conference service. I use GoToMeeting.
Leo: Oh, ok.
Greg: So if you want to talk to me I'll have a conference call with you but I won't talk to you in any other way.
Devindra: So nobody will ever call you because GoToMeeting, oh my God.
Alex: He seems lovable now, but if you call him at 2:00 AM, you get the beast.
Greg: Yea, you're going to get a fairly solid mouthful of invective. So coming back to 5G, there is part of 5G which is about standard. There's parts of it which is about getting 1GB to your handset, kind of pointless I suspect. You know, whatever. But there are emerging standards.
Leo: You know who wants 1GB to your handset is all the ad tech companies that want 1GB webpages.
Greg: Well what you want with the 1GB is you want the clock speed to be very high. Here's why. There's a secret here. If I can send data on a frequency of 1GB, then that frequency then is available for the next message. It's a thing called duty cycle. The faster I can get the message across the frequency and get it off—
Leo: Right. It opens up the channel.
Greg: Then the frequency's now available. So in some sense I want the clock speed or the bandwidth, the speed part of the bandwidth to be very high.
Leo: Just to keep the channel open.
Greg: Just to keep the channel open so I can multiplex more data across. But what we're also seeing is some people are thinking about can we use 5G for mobile broadband? So instead of using wide access for the—
Leo: Wait a minute. 5G's not mobile broadband?
Greg: There are group standards.
Leo: I'm confused.
Devindra: Sorry. Sort of similar to where there is that company from the Aereo guy sort of using millimeter wave technology, right, for—
Leo: So 5G is not a replacement for LTE? It's not the next generation?
Devindra: It's everything. It's all of these.
Greg: That's what I'm saying. There's part of 5G standards which is just balls to the walls speed. There's a part of 5G standards which is to be mobile broadband and widely shared amongst a small thing. But there's another part of 5G which includes small cells so in the current 3G what they wanted to be able to do is to take these little boxes and put them in your house. So if you had poor 3G reception in your house, you could deploy this small cell technology which would allow you to boost your signal inside of a 50 square meter, 100 square meter area. But that didn't work very well for 3G so in 5G they're building that into the standards. And now they're putting in—and there's another part of 5G which has to do with low power. You know we talked about NVIOT which is—
Devindra: Really, really low bandwidth, like kilobytes if that.
Greg: Yes. And this is where the low power comes from because it's literally on very low, low data volume. So low speed but they're using 800, the old analog TV channels, so the 800 MHz, 700MHz frequencies.
Devindra: And that's the most, honestly as much as we want, the fast speeds and the mobile broadband. That's the most fascinating thing to me. Like the idea that you have this really, really low power, even if it's a slow amount of data tech, that's going to be useful for all sorts of devices going forward.
Greg: Yea. So you've got this thing where you start to use different frequencies, or you're using different antenna designs. You're starting to see a lot of MIMO working out, beam forming technologies and—
Leo: So Susan Crawford who writes for Backchannel, she's a Harvard Law Professor says, "There is no— "she says it's hype. She says, "There is no 5G standard yet. ITU has not set a standard.
Devindra: Not until like 2018.
Leo: So it's like these other standards. It's a marketing term at this point.
Greg: No, there's a—it's a bit more than that. They're actually arguing over the standards.
Leo: So there is a conversation going on with ITU right now.
Greg: It's in the ITU. There is 5G. There's tens of millions of dollars being put into it.
Leo: She says, "for the moment "5G" refers to a handful of different kinds of technologies that are predicted, but not guaranteed, to emerge at some point in the next 3 to 7 years."
Greg: Yea, fair. Except I think—
Devindra: Well that's why we're talking about it. We're talking about all the different pieces.
Leo: I don't care. I'll be dead by then.
Greg: It's not going to be like 3G.
Alex: Don't be so morbid in the studio.
Leo: Well if they say 3-7, then we're talking 10 and you know.
Greg: No, no I think it will be a lot more fast.
Leo: You think it will be faster?
Greg: Yea what's happened is the big carriers have already stopped spending in the expectation that 5G. So they're winding back their capitol.
Leo: That's great. So our broadband today, our data rates today are going to go to crap.
Alex: Right, because their cutbacks are down.
Leo: Great because they have no money and they don't want to invest on anything because, well, it's all going to change in 3-7 years.
Greg: Yea, hmm mmm. So there's a whole bunch of competing things in here.
Alex: Keep building towers.
Devindra: Well a lot of them are prototyping things too like to get ready for the 5G wave. So I know that Verizon's like testing all sorts of like different antennas and different sort of like—
Leo: The rhetoric sounds so much like WiMAX. For instance, they're calling it Wireless Fiber, right, or Wireless Broadband. That sounds so much like WiMAX.
Greg: That's the high speed—so literally I said, 5G isn't one thing. Like 3G was fundamentally high speed data rates. That's it. HSDPA was the early versions of it, High Speed Data Packet Access. In 5G there's a part of it which is high speed data access and then there's low speed, low power data access. Then there's how do we do mobile broadband high density, low volume and spread it out over lots more areas, handling small cells, handling 800 MHz, the television frequencies. So as the analog channels, television channels are reclaimed, they can start allocating those frequencies to—
Leo: Is that part of the reclaiming of the low frequency bands? I think an auction is coming up.
Alex: There's an auction for that.
Greg: So if you can transmit a signal between 700 and 900 MHz, or maybe it's kilohertz now. But basically that signal, television signals go through walls.
Leo: 24 gigahertz goes nowhere. It bounces off a feather.
Greg: Yea (laughing). If you get enough people in a room, enough meat bags, it just dies.
Alex: Enough meat bags.
Greg: Yea. That's what wireless people call them, bags of meat.
Alex: Bags of water. That's really flattering.
Leo: You know the only thing wrong with our 5G deployment is all the meat bags.
Alex: If we got rid of the meat bags—so 5G is sort of—
Leo: I feel like this whole conversation with this and then the IOT things, it's all about getting rid of the humans so the computers can finally live.
Devindra: That's what we're building. We're building the ideal machine to machine interface.
Alex: Now who's the malware, Leo? If you don't get that joke, just wait.
Leo: Now who's the malware, baby?
Greg: You're talking, we're now talking that the maximum saturated market for people, smartphones, tablets, computers is around 40 billion devices or 2 per person.
Leo: ITU's saying, "Well we should have a standard by 2020."
Devindra: Sounds right.
Leo: Which is admittedly now, feels like a long time ago, but it's—a long time ahead, but it's only 4 years. God, that's weird, isn't it?
Greg: And there's somebody who said, like I have worked with companies that have written 10-year plans.
Leo: Right. So it's not unreasonable for them to think this far ahead.
Greg: No, and you're talking about one carrier can allocate a billion dollars' worth of budget in a 10-year cycle. So you know, and this is why Verizon's getting into content because 5G will disrupt their business models.
Leo: Ah, interesting.
Greg: So if you start to look at what Google—
Leo: You don't think that Verizon would be right at the leader of the pack in this?
Greg: Well, they've got a really good website on this.
Greg: It's heavily tilted in their direction. Nokia has got some very good information.
Leo: Why are they afraid of being disintermediated? Who are the companies that are going to beat them at 5G?
Greg: Anybody can start because small cells start to extend coverage and you can start to see municipal start to replace Wi-Fi with 5G.
Alex: So will this help municipal broadband efforts then?
Greg: Yes, potentially.
Alex: That's amazing.
Greg: No last mile.
Alex: That's so huge.
Greg: So America's got a—in terms of regulatory control and infrastructure that's in place, America's heavily burdened with its coaxial cables, right? So they, the government said, "If you run these cables we'll give you a monopoly over the extracting profits from the cables that are in the ground." So Comcast—
Leo: Right. And that works so well. Yes.
Greg: Now if we can bypass those cables in the ground and start to deliver data volumes over—
Leo: At a fiber-like rate.
Greg: At a DSL-like rate.
Leo: DSL-like rate.
Greg: Let's start there. Let's be modest.
Leo: I'll take a megabit.
Greg: Yea, a megabit per second and all of a sudden you're starting to, your universal service application starts to go out the door. You don't need to run copper into houses because everybody's got mobile. Cable, having coax cable in the ground means I can start to bypass government mandates.
Leo: So this is why that dumb pipe guys are anxious to get out of the dumb pipe business. They see no future in it ultimately. At least the hardwired pipes.
Greg: Yes. They should—but you're looking at exactly the same business models that went on with electricity.
Leo: Is this why Verizon stopped laying fiber? They gave up on FiOS because they knew that ultimately it wouldn't be enough, that this wasn't a long term?
Greg: So they're still laying fiber but they're laying it in the back hall.
Leo: That's where you need it. The back hall is the connection from the central office to the cell tower. You need the speed of the cell tower. And after that, there's no last mile problems.
Alex: What does this mean for Google Fiber? Are there any implications for how Google could roll out its own nationwide—
Leo: Is Google building their own fiber plants or are they just taking over?
Alex: Talk to the guy over there.
Leo: You don't know?
Greg: They buy fiber from the manufacturer. You don't make fiber.
Leo: You don't build your own. So not make it but I mean lay it. Are they laying their own fiber? Are they digging the trenches? They are.
Greg: They are digging the trenches and putting it in.
Leo: Seems like that is the business that you would not like to be in.
Greg: You would think that potentially we could start to use 5G. It's a 5-year horizon, a 5-10-year horizon. But when you're building, burying fiber into the ground, you're looking for a 20 year ROI cycle, right? It doesn't come back. It's like $500-dollars a meter.
Greg: When you take into account trenching, safety—
Leo: So the answer to your question is, this isn't a good idea for Google to be doing at this point.
Greg: Oh no, it's absolutely a great idea for Google because it forces the carriers to change their productization. Comcast and various other carriers have monopoly positions and—
Leo: They need, Google needs to push the future forward, and if they can do it by using an old technology to scare these guys, they're going to do it.
Greg: And they're only investing modest sums.
Leo: That's interesting.
Alex: Well, we've seen the Google other bets returns in their SCC filings. It's very expensive. In aggregate, all their other bets, their moon shots.
Greg: And also don't forget that big companies are just stupid. So Verizon might well have just, you know, rolled out some fiber and they decided to spend it on upgrading the executive's offices in New York or something instead. And then if nobody was looking, we'll just keep that money.
Leo: Crawford says because of the bag of water problem, you're going to have to do backhauls not to a cell tower, but almost right to every single home.
Greg: And that's exactly what there are versions of 5G standards being promoted to do. Point to point links and there's also standards where the towers can talk directly, they can actually form backhaul mesh networks between them. But these are all things. These are all arguing in the standards bodies, right?
Leo: What's hysterical is we are about to elect a new president and a congress that thinks that Facebook is the state of the art in internet.
Alex: Trump does email.
Leo: I feel like the real problem here is that regulators are completely out to lunch on this stuff. They don't understand it very well. Maybe the FCC does.
Greg: I think they do understand it but they operate—and keep in mind that we need to be careful to separate US telecommunications law from European telecommunications law as we do from what happens in Asia or in the middle east or various Oceana, Australia. Australia has a different problem, a similar problem to America, very large land mass. The physical land mass of Australia is the same size as mainland USA. So if you want to start running bandwidth—
Leo: Population density is still only along the coast.
Greg: Yes. But there's also a government monopoly that one of the biggest employers of the country is the National—they don't want to substract jobs. In the US it's the bargaining power of the TelCos in Washington to influence things to go their way. And in Europe it's a free for all. It's very much a community driven thing. So you're probably going to see the innovation come out of Europe.
Leo: Crawford says, "So a competitive, terrific 5G future looks great in places like Tokyo, where you can get dark fiber capacity anywhere in the city for about $50 a month. It looks good in Stockholm, where Ericsson is teaming up with SK Telecom to test 5G applications. In the US we're going to be fundamentally behind in 5G."
Devindra: That's not new, yea.
Leo: So once again.
Greg: Innovation's going to come from somewhere else and then once it proves out, it will come most likely and that's due to the way that telecommunications regulation operates here in the US compared to other markets.
Leo: There are some technical hurdles, but more importantly, there's some just physical hurdles of just getting that high speed data close enough to where it needs to be.
Greg: Yes, so when you have, if you've got signals being transmitted at 700 MHz like the analog television channels are, it goes through walls. They go for miles. You can make them really powerful or really weak and they'll go very far. I feel transmitting, if you're trying to get a gigahertz, you know, like a gigabit per second, you need to be transmitting at 24 gigabits right up at the top of the spectrum, 24 GHz.
Leo: 24 GHz, that's the bag of water right there, boy.
Greg: So then you, that might be suitable for backhaul, so from mobile tower to mobile tower, or cell to cell.
Leo: How about the Galaxy Note 7? What do you think of that (laughing)? That ain't 5G. Have you tried the eye reader?
Devindra: Actually I had brought it here to the hotel with me because I'm going to be writing it up. It's cool. It's a little more comfortable than the Gear VR.
Leo: So the 950 and the 950XL also had an iris scanner. And it did not—I remember you had to go up, back, back, no, no, no, like that. Is this better than that?
Devindra: From what I've seen it is. I thought you were talking about the Gear VR not the iris scanner.
Leo: Oh, the Gear VR, that's cool too, yea. Yea, I have the new camera too. I bought the 360.
Devindra: But yea, it's also like with a Note, the whole point of it was a very big screen on a smartphone.
Leo: Everything has a big screen now.
Devindra: So if the S7 Edge—
Leo: People like you, bully that you are, Alex Wilhelm, mocked me when I held a giant Galaxy Note, the 1st Note to my head.
Devindra: People are always holding them up to their head.
Alex: You look silly.
Leo: But it's the same size as this phone. So we're now used to it, right?
Devindra: We're a little more used to it. I think the S7 Edge is a 5.5" screen. The Note 7 is 5.7." Everything is a tablet more.
Leo: It's a little bigger, yea.
Devindra: I like smaller phones still.
Alex: See that's interesting because back when, before the iPhone 6 came out, iPhone and Apple fans were saying things like, "Android phones are too big. Are they tablets? Ha, ha, ha." And then the iPhone 6 and 6S came out. And they were like, "Oh, this is great." 6 Plus, sorry.
Leo: Because Samsung's Note was selling really, really well.
Alex: People like more screen space on their mobile device. It's their main computer to your point about Chromebooks.
Leo: We can thank Samsung really for making the big phone.
Devindra: I think the iPhone 6 Plus—those are among the worst designed iPhones I've ever used. They're top heavy in a way that isn't great and because of the way Apple design works, so I have a broken 6S with me right now.
Leo: You broke your first phone.
Devindra: Not my first one. But there's a lot of bezel.
Leo: Yea, way too much bezel.
Devindra: On the top and the bottom of iPhone screens. And because you have that on the Plus, there's much more space. Like it's a much bigger device than it needs to be. You can hold an S7 or even a Note and it feels a little—
Leo: It feels smaller, yea, yea, yea.
Devindra: It feels smaller. So I think that Apple hasn't really—they didn't have a way to master that.
Leo: This feels smaller.
Devindra: That feels smaller. It's a great phone. So I'm not a big fan of the Plus. I know people like it. I remember just trying to walk down the street and use it during my review, just couldn't do it. And that's something I have to do a lot like check on email or type in something in my apps.
Leo: Like Pokémon Go.
Devindra: Whatever. Pokémon Go is important. But even if I'm at home just like reading in bed or something, that's a phone that would often just like plop right on me. And it's like—
Leo: Are you unusually weak in your forearms?
Devindra: No, it's a wrist thing. It would be more wrist but even then it's just like not great.
Leo: Really? Not great?
Alex: I've actually never dropped my phone sober when I had it in my hand.
Leo: Well that is the problem. So all right, we've got to bring it up then. Apple is about to really, I think, and it's just a few weeks away, with a new iPhone. Launch into kind of unknown territory. They're going to remove the headphone jack. And it's really—the world is divided among those who say, "Finally, they're getting rid of that antiquated, analog technology."
Devindra: There's a handful of people that are saying that.
Leo: Well, somebody's saying it. Jon Gruber's saying it.
Devindra: You mean he's defending an Apple position?
Leo: He's defending an Apple position. Is Apple heading into a buzz saw with this one?
Devindra: This has been something I've been fighting for the past couple of months because being a gadget blog we've been hearing this for a while. I've seen the Moto Z and the new—
Leo: That also don't have a headphone jack.
Devindra: And the Moto Z feels really thin and crazy thin and I guess I can see why.
Leo: Does it?
Devindra: Yea. The other one, the Moto Z Charge I believe, the slightly more powerful one, that one is the full size of an iPhone. Still has the headphone jack and it is just baffling to me because I can't think of a single port that is as universally accepted that works on so many things.
Leo: We all have so many headphones.
Devindra: We all have so many devices that work with it. It's like the one thing that you know when you plug something into it, you know what's going to happen. Let's say you're trying to get a microphone in there or something. That's a little different.
Allyn: There are things even like credit card readers.
Leo: I'm sure Square's going to make a lightning version.
Allyn: But it will cost more.
Alex: And it will be less reliable than the old one. I use Bluetooth headphones when I'm running and they're great, right? They're fantastic. But they're not as good. The fidelity goes down, they're less reliable. I'm giving up flexibility.
Leo: I think a lot of people are saying Apple's making a decision in favor of wireless headphones. I don't think that's the case.
Devindra: It's probably straight to lightning.
Leo: I think they want lightning headphones.
Alex: Oh, I really don't want that.
Leo: It will only work with an Apple device. It won't even work with an Apple laptop. That's the irony of it. Apple's still putting and I imagine will continue to put headphone jacks in their laptops.
Devindra: And even if they put in a dongle which this is what a lot of people are saying, that's not a good solution.
Alex: Right, because the current MacBook doesn't have any other ports besides USB-C or whatever it is. Ok, so—
Leo: So here's some heresy. Given that Apple has not really kept up with the technology, they're not making a phone that is in any respect the equal of most of the other high-end phones coming out form other manufactures, they're going to eliminate a port that everybody wants and expects, is it possible this iPhone 7 could be a real flop for Apple?
Devindra: I mean it's going to be, it's going to be an interesting to watch. We forget about was it the 5C? Or which?
Leo: 5 SC.
Devindra: No, not SC. The iPhone 6C? The one that was C, the colored one.
Leo: The colored one, yea. It was a flop.
Devindra: It came and went. No one cared. This is a main model and it's going to be a bigger difference. Like this is, because of what I do, I've been buying a new iPhone every year. I know that's not a normal thing. This year it's like, the 6S is a great phone. I'm not going to buy—I don't think I'm going to buy any phone that has no headphone jack. Why would you do that?
Alex: Well, you wouldn't. But Leo's point of it being a flop, I think we need to keep flop in the correct context because it will sell very well. But I don't think it will help Apple drive iPhone unit growth. So it will be a flop in that it won't actually accelerate—
Leo: Will it sell very well?
Alex: I think it will sell as well as any iPhone.
Leo: You don't think people will turn away from it?
Alex: I'm not going to.
Leo: Will this not make Apple users start to look at other brands?
Devindra: Maybe a handful but—
Leo: No, they're loyal. They're really loyal.
Devindra: Either that or—
Leo: You'll buy one even without a headphone port?
Alex: I'm going—the home screen on my current iPhone is basically the same one I've had since my original iPhone, since 2007. It's like 9 apps.
Leo: It absolutely is because Apple has not changed screens appreciatively.
Alex: And you know what? It works. And I like it. And I don't really want more stuff. There's nothing on Android that they can do that my iPhone can't that I'm desperate to have.
Leo: Well that's my point, that the headphone jack may be that thing.
Alex: But I use wireless headphones now. So I feel like I've left the discussion because I feel weird. But last time I was on TWiT we had the same talk about headphone jacks and so forth and I was on the team that believed it should be kept. I think they should still exist but I don't really have the direct use case now. So maybe—
Leo: Bleak in our chatroom says, he splits the difference as well. He says, "Apple's doing the right thing, but they're going to pay a price initially."
Devindra: Everyone says it's the right thing, right? We've got to kill this analog hole because reasons.
Alex: It's the floppy example, right? They dropped floppy ports.
Devindra: And it's not accurate.
Alex: Well, people say you shouldn't argue with because they were right about floppy drives.
Devindra: They were right about CD drives and whatever. And that's not necessarily true because it's a completely different format and just the way they're handling it. Like there is no benefit to direct digital connections. Like you may get slightly better DACs but—
Greg: What if it's a mag safe concept?
Devindra: That would not be great for headphones. You want headphones to stick in.
Greg: The problem that you have now is that if you're not buying and MFI lighting connector—
Devindra: That too.
Greg: Then it rips out. Actually people are starting to damage their lightning connectors because the point of a lightning connector is its meant to smash, like destroy itself if it goes to a certain—part of it is the mechanicals, right? Nobody ever thinks about connectors as mechanicals. There's a certain pulling pressure which it's supposed to pop out. And if you put like torsion to it, it should snap off so that you don't damage the phone. And that's part of what's MFI.
Devindra: You want to take a look at my phone, it has it's basic—
Leo: MFI's made for Apple. It's a certification program. It also costs companies license fees.
Greg: Yea, it's 75 cents per connector which makes it very expensive. It's not framed for any reasonable and nondiscriminatory licensing. I think that the trick here is that Apple—I think the real question here is the value for money. I think Apple, my perception of Apple, and I'm an all time—like I have a Watch, I have a MacBook, I have an Apple iPhone 6 and I use my classic iPod. So I do not listen to Apple Music. I listen to it on my classic iPod because that's still the best medium for me. If Apple releases an iPhone this year, I will very seriously question whether I'll by one, not because I don't need a new one. I do because there's a hierarchal plan in my family where they get passed down. But I don't think that Apple products are good value for money anymore. If I'm being asked to spend $700 US dollars or $700 British pounds on an iPhone, I'm looking at that and I'm going, "That's not—I'm not getting that." You know, back in 2007 when the first iPhone came out—
Leo: That was value for money.
Greg: That was value for money. New cameras in 2012, 2013, there's real value for money in those products.
Leo: Well this is my question. This is suddenly a crack in the Apple façade that could end up being very damaging to Apple. If somebody like you is considering not buying an iPhone, I mean for me the tipping point was the Mac Pro, the trash can Mac which I frankly feel like I was robbed. I spent 3 or 4 thousand dollars on that.
Devindra: Although we've seen how you handle hardware, Leo. Like you took a look at this computer and you were like, "Poof. Purchasing page. Let's check this out."
Leo: Well I do buy a lot of hardware. That's kind of my job. But I wanted the Pro to be my workhorse, right? And it's not upgradable. Apple hasn't done anything to it in 2.5 years.
Devindra: It's running slower video cards, yea.
Greg: But the thing about the Mac—
Leo: That was supposed to be their best Macintosh.
Greg: That's right. And the trick about the MacBook Pro is that the GPU that's in it is still the most modern server CPU. It's—there has not been a—
Leo: There hasn't been much improved since then.
Greg: And that motherboard is highly integrated. And what you really want is maybe a better GPU.
Leo: I wish I could upgrade the GPU.
Greg: Yep, and so on and so forth. But in reality, there's actually been no advances from Intel in that technology.
Leo: And some have said, in fact this is the article we referred to last week on Screen Tech, where the problem is really they said is Intel hasn't improved.
Greg: Yea, I subscribe to this article. I subscribe to the ideas put out in this article is that things haven't—but my point is, you know the rumors are that there is a new Apple Watch coming. Are you going to pay $600 dollars for an Apple Watch or $900 dollars for an Apple Watch?
Greg: I bought an Apple Watch off eBay second hand.
Leo: Looks good. Looks like mine. I spent a lot more for it than I bet you did.
Greg: I bought it second hand about half the price that you pay for new.
Leo: It's a pretty brisk market actually for used Apple Watches I would guess.
Greg: And it's not worth it. I would not recommend at the price that I paid for it, at half price, it was not worth it.
Leo: Meanwhile, Apple's spending apparently a lot of money and energy on something like an Apple Car.
Leo: Should they be spending more time on their existing product line and less time on the future? I mean you've got to plan for the future.
Alex: I style like Tim Cook.
Greg: So here's my take on that just for what it's worth and then someone else should have a go but my belief is that Apple's building and operating system for cars, not a car.
Leo: Right. That's the latest thought as well.
Greg: But they haven't been very successful with that in General Motors and Ford have basically said, "We're going to."
Leo: It's directly contrary to the reason Apple exists which is if you want to have a great hardware, you have to make the software that runs on it. To decouple software and hardware—
Devindra: It's going to be the Windows for cars.
Leo: And in fact, even somebody like Elon Musk has said, "You really need to make the hardware and software together because the lessons you learn from the hardware form the software." To decouple them is probably not the best procedure.
Greg: Apple has to have a battery plant.
Greg: So Tesla has taken over these buildings.
Leo: Sorry, he's got the Gigafactory, sorry.
Greg: And China's got Gigafactories too, but they've got battery manufactures. So are you telling me that Apple's going to own a car?
Leo: I thought the whole idea was crazy but again, I would say they should be focused on the—I think maybe it's time to put a little more energy into the practical.
Alex: Do you—oh, sorry, go for it.
Leo: Let's take a look at this Ultrabook you've got here. This is a Lenovo and it's an OLED screen. This is their ThinkPad X1 Yoga. So it does yoga, right? It does downward dog? It does warrior pose? There's your downward dog, yea. That OLED screen though.
Alex: That's really nice.
Leo: Do you like it?
Alex: If you can't tell on the video recording, it's just gorgeous.
Devindra: It's just like—
Leo: I love my OLED TV.
Devindra: OLED is the future of like display. That's what I want everything to look like.
Leo: In fact, Apple's even kind of fallen behind with its IPS displays on the iPhone. Rumor is next year they'll be doing an OLED.
Devindra: Next year, well next year—
Leo: But Samsung makes all the screens right?
Devindra: Samsung and LG.
Leo: LG but between the two of them, I hope Apple can get a few screens.
Devindra: What are they going to do. Wrapping up like the whole iPhone discussion, like next years model will probably be the one that gives you the value you're looking for, for maybe giving up a headphone jack.
Leo: Typically, that's what they've done in the past, tick tock. This is going to be a tick, tick, tock kind of.
Devindra: Sure. So it may be a disappointment but—
Alex: No, I'm good, it was a great little phrase. I liked it. It's a tick, tick, tock.
Devindra: That reminds me of when the 4S happened, right? That was the first, let's do that design again with slightly better hardware and now we're doing it for a 2nd time.
Leo: We haven't done that 3 years in a row yet.
Greg: I don't think we can keep expecting hardware to double every 2 years.
Leo: No because we've reached peak phone. We kind of know what a phone should be.
Devindra: That's what, $300 dollars, that Axon?
Leo: $400 dollars for this Axon.
Devindra: For a premium looking phone.
Leo: In every respect. Dual front facing speakers with Dolby Atmos Sound though that means on a phone—
Devindra: It's just software.
Alex: I think you know we finished a product categorically when you've ground all the profit margin out of it across the world. Computers, televisions, phones.
Leo: But this phone, to be fair to Apple, would not exist if Apple hadn't invented all these products and created a manufacturing process in China that could crank them out at huge volumes.
Devindra: We wouldn't have Google Glass without the iPhone.
Leo: They're benefitting, they're benefitting from the fact that these sensors, the radios, everything is cranked out in mass volume at very, very low cost now. The problem is Apple isn't dropping its price to match the lower cost and others are.
Alex: Well you have the 5SE which is designed to have a lower ASP.
Leo: $100 bucks lower.
Alex: Which matters.
Leo: I wouldn't say it doesn't matter but it's not enough is what I'm saying.
Alex: No, I agree with that entirely. It was a good attempt, it just flopped.
Greg: So I think you're going to see Apple's average selling price to drop. Whether it will drop this year or not and this comes back to the headphones thing. I think with the value for money dropping the way it is, and if they suddenly switch headphones then the backlash could be enough to have an impact on the bottom line. And Apple's a hugely profitable company and they might not even notice. You know, I mean they're already using TRES instead of TRS on the audio jacks, you know the 4-way jack. They don't have standards, the more modern standard for it. It's why they have that audio, you know when you have those little buttons and you press for—
Leo: For the microphone.
Greg: For the microphone, right. And it's typically a TRS jack. If you move to Bluetooth, you know, I don't own a pair of Bluetooth headsets. I have no idea why I would buy any because I already have got enough stuff to charge.
Alex: Oh, I bought mine for like $18 bucks and I broke my iPhone's little port thing we're talking about, so you know, it worked out. I had to get a new phone, so.
Leo: I mentioned this before but Steve Gibson has a theory that Apple will do noise cancelling headphones because with the lightning jack you have power and you have computing capability so you could actually in theory make these headphones something compelling. If they do that, that's different. You have to give me value for talking away the headphone jack.
Allyn: They'd have to put those in the box though.
Leo: They're not going to do it by the way. I know they're not going to do it. It's just a theory.
Greg: What about DRM? The lightning connector—
Leo: That's the other story. Cory Doctorow wrote on Boing Boing this week, and of course Cory is a very strong advocate against DRM that that's all this really is.
Greg: I'm really concerned just about I've got all sorts of audio accessories—
Leo: Close the analog hole.
Greg: Yea, the lightning port doesn't work. The TRS jack, the 3.5mm audio jack works sweet. And I'm yea.
Devindra: Are we still looking for stories?
Leo: Always but let's take a break, and then we'll get some more stories. Please, guys, find the stuff you want. We can talk about iMessage and Matthew Green the cryptographer for Johns Hopkins saying, "You know, it isn't really end to end encryption. It's not much stronger than SSL. In fact, if you don't, if messages aren't delivered, anybody can see them, you can go to Apple and say, ‘What do these say?' it's really not end to end encryption." We can talk about that. There's lots to talk about. But before we do that, we should talk about your favorite way to make a phone call, Greg Ferro. You didn't know this, but GoToMeeting is a sponsor.
Greg: Did I segue again?
Leo: You did it again. You did it again. You're a natural. You should have a podcast. GoToMeeting is the online meeting tool that makes you, it's so easy and so quick to make more than a phone call. You can start with a conference call. In fact, GoToMeeting you just click a link, it sends an email out and you can have a conference call to do that. But then, if you want, turn on cameras. Now you're seeing your client face-to-face. You can turn on screen sharing. Now you're seeing their document or they're seeing yours. You show them your PowerPoint. You collaborate together on documents. I've actually, GoToMeeting's so fast and easy to use, I've used it to prepare a speech with somebody that we were going to give jointly. We were in different places in the country and we went over it together, even the timing and everything with GoToMeeting. You can send private chats and video links to your meeting. You can pass off presenter duties with ease. And it's free for 30 days. There's no reason not to become a meeting MVP no matter where you are with GoToMeeting. Go right now to GoToMeeting.com. Click the Try it Free button and from now on, instead of starting a phone call, start a GoToMeeting and you'll have all those capabilities available to you. GoToMeeting.com.
Greg: I can really highly recommend it. I've used all the different conferencing platforms.
Leo: This is the easiest. It's easy for your clients because they get a link, they click it. Even if they've never installed it, it installs really—kind of like 30 seconds and they're there and it's running.
Greg: It's real easy to use USB headsets and get high quality audio like you were talking about VOLT and high quality audio. A simple point of fact is that your iPhone is not designed for good voice quality.
Alex: I'm shocked.
Greg: It's designed for data. And so if you have a headset, then your GoToMeeting becomes way more—
Leo: Because it's data.
Greg: Because it's data. Yea, so, I can highly recommend it.
Leo: When they first created the telephone, as I remember the calculation was, because of course in the early days it was wired point-to-point, you wanted to use as little bandwidth, whatever that is bandwidth as possible. What is the lowest bit rate you can use and still be intelligible? And it was something like 6 bits per second.
Greg: 5.4 bits.
Leo: 5.4. And it still is.
Leo: (laughing) the whole thing has changed but it's like movies are still 24 frames a second. It's to save film stock but we just, I guess we got used to it. We expected—now when you get a real phone call, a real voice, it's like whoa. That's why we were so thrilled when Skype became widely available. That's what makes this network possible. Because we could do a podcast with people all over the world.
Greg: You always know a mobile phone call because usually the compander's running at about 5.4 kilobytes per second. And that's why it sounds so muddy. They've taken 400 hertz and compressed it.
Leo: And you don't really need to. Not anymore.
Greg: No, they can easily run a 32 bit—
Leo: You could have an 8-bit CODEC. It's crazy.
Greg: Yea, that's why you use GoToMeeting for everything.
Leo: I don't blame you. We had a wonderful week as you may have known if you saw the shows this week on TWiT. But if you didn't, watch this fine tiny little movie all about it.
Narrator: Previously on TWiT.
Dick DeBartolo: John likes to label things. He even had a label on his blank.
Stacey Spears: John's label maker has a label.
Narrator: Know How.
Mark Smith: In a world where Know How talks about how to build a DefCon badge.
Father Robert Ballecer, SJ: And we thought it would be interesting to assemble one of these DC DefCon Dark net badges that you talked about.
Mark: I told you man, you could do it. It just took a few years.
Leo: Our guess, I'm really glad to have him, Paul Adams. You worked at Google + early on, right?
Paul Adams: I did, yea. I was there from its inception.
Leo: Loved Google +. What went wrong?
Paul: You have a dingy nightclub and it's like happening in there and people were having the time of their lives and it's full. And someone opens a new nightclub next door, the edges are better but it's basically the same. You're just not moving everyone in next door.
Narrator: This Week in Law.
Denise Howell: The moral machine is a platform for gathering human perspective on moral decisions made by machine intelligence, shows you moral dilemmas where a driverless car must choose the lesser of two evils such as killing two passengers or 5 pedestrians.
Amanda Levendowski: If you tell an autonomous vehicle that anyone crossing against the light is fair game, New York City is in real trouble if they start rolling out autonomous vehicles.
Narrator: TWiT. Making the world safe for technology.
Leo: And if you thought that was good, well, Jason Howell has a look at the week ahead.
Jason Howell: Thanks, Leo. Here are a few stories we're going to keep a close eye on in the week ahead. First, apparently, Minecraft is coming to the Oculus Rift sometime this upcoming week so we'll look forward to that. Also if you happened to pre-order Blackberry's DTEK50 Android device, they start shipping on Monday, August 15 though I'm guessing you probably didn't do that. The Intel Developer Forum happens Tuesday, August 16th through 18th in San Francisco. You can usually expect some announcements to come out of that event. On Wednesday, August 17th, you too can drop $5,000 dollars on GoPro 6 camera Omni Virtual Reality recording rig. No big deal. On Friday, August 19th, Samsung's eagerly awaited Galaxy Note 7 is officially available on all major US carriers though it will cost you a smooth $850 dollars if you buy it unlocked. Also on the 19th, Samsung is releasing its Gear 360 VR camera for $350 dollars in the US. That's only $4,650 dollars less than the GoPro rig. Just saying. That's a little look at the week ahead. Back to you, Leo.
Leo: Thank you, Jason Howell. We'll be catching all of that every Monday – Friday, 4:00 PM Pacific, 7:00 PM Eastern, 2300 UTC actually on TNT but actually I think we're not doing a Friday—are we not doing Friday's TNT? Pardon me? It will be at 10:00 AM on Friday to give them time to tear down the studio for the final time. So tune in a little early on Friday for TNT. All right. You all get to pick a story that we didn't cover that you wanted to cover that you're mad that Leo didn't even think of it. Now they're all working furiously but I'll start with you because I know you have one, Devindra.
Devindra: I've got one ready to go, this week's Snapchat had another face mask issue, another racist issue with the mask.
Leo: Oh, so the last one was to celebrate Bob Marley.
Devindra: The Bob Marley one.
Leo: And it made, it put blackface on you. Isn't that bad?
Devindra: It's pretty bad.
Alex: It's pretty bad.
Devindra: Even if it was supported by the Marley Estate, you don't want to do that.
Leo: It's pretty bad. I mean I agree, blackface is bad. And it was a big thing in college campuses over the last couple of Halloweens to do blackface.
Devindra: I remember in college I had to explain this.
Leo: It's terrible.
Devindra: That's bad. Yea, don't do that.
Leo: So this filter was bad. So now they're not doing blackface anymore.
Alex: Now they're doing—
Leo: Yellow face.
Devindra: Yellow face which is not great.
Leo: Oh, that is bad.
Leo: Oh, I hadn't seen it.
Alex: Oh, it's bad.
Leo: It gives you slanty eyes and buck teeth.
Devindra: It's not great. It is not great at all.
Leo: That is the racist imagery that was used during World War II.
Devindra: And they're saying that, oh they were trying to go for anime imagery. Anime doesn't have buck teeth, dudes. That's not how it works.
Leo: No, they have big eyes.
Devindra: So yea, this is—
Leo: That's blatantly racist.
Devindra: Everybody complained. I'm more like, "Why the hell—how is it that you've already done this?"
Leo: Don't you know?
Devindra: Be smarter. Be better. Come on.
Leo: But you had a good reason for this.
Devindra: I had a good reason?
Devindra: Oh, yea.
Devindra: It's mostly programmers. It's probably you know—
Leo: It's a bunch of white, young guys, privileged guys. Hey, and they don't have—like you, Alex, and they don't have that sensitivity. Me too, to be honest.
Alex: Well, the young part.
Leo: Old white privileged.
Alex; There you go.
Devindra: But this is what happens when you don't have enough diversity in your culture and somebody would walk by and say, "Hey, don't do that. That's bad."
Alex: Right, because everyone at this table was instantly like, "Oh, no, terrible idea." This made it through ideation development and testing.
Alex: And PR and launch. And they they're like, "Oh, crap."
Leo: I would argue that this is worse than the blackface one.
Devindra: This is probably worse.
Leo: This is really bad.
Devindra: I think they left the Bob Marley one up for a while. They took this one down immediately. But this one is, they're both pretty bad.
Leo: They're both bad.
Devindra: They're not good. And it's more like—for a company that is so young and growing so quickly and valued so much, it's astounding to me that they can be this tone deaf because representation matters.
Leo: Uber did the same thing. They seemed to have learned not to do that stuff. They probably just buried the sentiments and made sure they don't come out. There seems to be this kind of thing going on and now, that's the best argument I've heard yet for diversifying the workforce in Silicon Valley. It's not just because it's the right thing and the fair thing to do. You actually are making worse products because you are a monoculture.
Devindra: It doesn't reflect the people who are actually using your products, too. So.
Alex: It's like when people argue that car ownership is dead because of Uber. Like today, we almost didn't get to come to this show because our Uber drivers wouldn't drive us to Petaluma.
Devindra: We got in the first Uber, and we're like, "Thanks for taking us to Petaluma." They're like, "Nope."
Alex: "Get out." So we tried again.
Leo: But when you call an Uber, you tell them we're you're going.
Allyn: The driver doesn't know until he gets to where you are.
Leo: Oh, he doesn't.
Alex: I got into two different cars, pulled up and then told to get out. And then we took a regular taxi all the way here.
Devindra: It was crazy expensive but he got us here, so.
Leo: That's all right. We'll pay the fare.
Devindra: It's all expensed. It's all expensed.
Leo: We're glad to pay.
Alex: The point is that people—wait, wait, wait. Just wanted to point out that Uber's no threat.
Greg: I'd just like to point out that I had no problems getting here. Maybe it's the people.
Allyn: Then you got a good Uber driver.
Alex: That's actually probably not all wrong.
Leo: Yes, you look like the type that's going to throw up in my car.
Alex: This iced tea's killing me. But you're pointing out monoculture and it's negative. People say that you can get around now without a car because of Uber. But what if you have a disability or you have a dog or you have a child.
Leo: Or you're normal and you don't use Uber. You don't have a smartphone. There's all sorts of reasons.
Alex: Right. And so I think to your point, it's very accurate that if we had a more diverse workforce in Silicon Valley, it will lead to not just different prospects, but better products for all of us. And in this case, I can do the smell test by how there is universal condemnation of Snapchat and almost no, that I saw, response that it wasn't actually bad. And that proves that they were so far over the line, that no one even got upset about it.
Leo: Nobody defended it.
Alex: No one was like, "Oh, yea, Snapchat."
Leo: Even Milo Yiannopoulos did not defend.
Alex: Don't, don't, don't, don't do that. Don't bring him up.
Devindra: Don't even say the name.
Leo: It's like saying Betelgeuse once. You only have to say it once.
Alex: I'd rather hangout with Voldemort.
Leo: That is actually appalling. I am shocked.
Alex: I have a different story.
Leo: Did they apologize?
Devindra: Their excuse was that they were going for anime style. I don't know if they actually apologized.
Leo: That is not—there is no excuse. You just say, "Oh, God, that's horrible. You're right. Sorry. We're taking it right down." There's no—you can't say, "We were going for anime."
Alex: Even if they were.
Leo: Ok, your story, Alex Wilhelm.
Alex: So my story is a headline from The Register that I'm super excited about. It raised a kerfuffle between Adblock Plus and Facebook. And I'm going to do my best to read this.
Leo: We need to talk about this, yes. Let's talk about it. Soo Adblock Plus says we're going to block Facebook ads. Facebook immediately applied a technology to block Adblock Plus. Well, Dan pointed out that Adblock Plus in their fix of that ended up deleting people's posts as well as ads.
Leo: I want to know where it stands now. Is Adblock Plus on top or Facebook?
Alex: Well, listen to the headline and tell me if you can understand where we currently are. All right, so, Adblock Plus blocks Facebook block of Adblock Plus block of Facebook block of Adblock Plus block of Facebook ads.
Alex: This arms race will never end.
Devindra: I wish I could have written that.
Leo: That's a really great line (laughing).
Alex: So, Adblock plus blocks Facebook block of Adblock Plus block of Facebook block of Adblock Plus block of Facebook ads.
Alex: In other words, I don't actually know what that means to answer your question but I thought it was a great headline.
Devindra: It was really hard writing about this story just because Adblock's name is Adblock and then it's about ad blockers and then it's about Facebook blocking them. So, yea, I could understand that.
Leo: Well and Engadget is ad supported and a lot of pages are ad supported. We're ad supported.
Devindra: Well, the web is.
Leo: And I have to say, ironically Facebook is not the worst by any means, offender. Their ads are benign compared to most pages.
Devindra: Usually pretty good, yea.
Leo: Yea. I mean they're on right there. I guess they have—
Greg: I guess they haven't bribed Adblock Plus to get on the—
Devindra: That to.
Leo: Oh, they're not acceptable ads.
Greg: Specifically, Adblock Plus makes its money by taking fees and proof that they are suitable advertiser or trusted advertiser so that Adblock plus lets their ads though. So if you install Adblock Plus, you still see Google ads and various ad networks because they guarantee that they're putting clean ads. You can buy your way past the ad blockers.
Leo: You can turn off the acceptable ads policy on Adblock Plus for users so you don't have to see those ads. Secondly, those ads still have to adhere to the standards, the acceptable ad standards which are set by a larger group than Adblock Plus and are used by other ad blockers.
Alex: Why doesn't Facebook just buy Adblock Plus, turn Adblock Plus off of Facebook ads and then block all other ads on the internet.
Devindra: You figured it out.
Alex: Think about it. They should do that. It can't be very expensive. Facebook has a lot of money.
Leo: Just block all the other ads.
Alex: In fact they could have bought them, done that and survived.
Devindra: At that point the FCC is probably like, "Hey, no, let's not do that."
Alex: Then buy it and don't tell anyone.
Greg: Then you're interfering with people's personal preferences and you're getting very close to American Constitutional process.
Leo: That's all right, we're already there. Let's just go all the way.
Devindra: Have you seen America today?
Greg: Adblock Plus looks for HTML tags and java script tags inside of the code and as fast as Adblock Plus works out the patent matching up that they need to do, Facebook then iterates their webpage and does a constant, instant deployment.
Leo: I'm curious because the simplest way to defeat this is by being a first part ad. The reason that these ad blockers work is because they're 3rd party ads coming from a source other than Facebook.com. Is that not true?
Greg: Not always. So Facebook has a lot of its own ads now that it generates.
Leo: Would those be blocked?
Greg: So you have to look for the tags in the HTML code.
Leo: So they are looking at now deeper than just the source. They're looking at the actual content. They're doing a deep inspection.
Greg: doubleclick.net is where a lot of Google's ads DFP program comes from and you know, there's like a thousand—
Leo: So first party ads could be blocked as well now. That's interesting.
Greg: Well that's what we want to be able to do that. But you have to be able to recognize—
Leo: Well I don't want to be able to do that.
Greg: There has to be a consistent tag mentioned, a HTML tag to work. And so Facebook just keeps iterating its code.
Leo: We had a guy on yesterday on The New Screen Savers proposing a new form of ad blocking. The site is ourdata.us. Friend of mine, David Glickman who was a multiple, serial entrepreneur, he said, "Well what if we did it this way? We have an ad blocker. Runs under iOS with the official ad blocker check mark and all that stuff. And instead of blocking ads, we go to content companies and we get them— "let me see if I can get this right. "We get them to pay us and then we pass the money on to you." In other words, you get paid for your attention. Otherwise you don't see the ad. Because the real problem here is, and I'm sure this is a problem for you, it's a problem for PC Perspective, it's a problem for Engadget, I doubt it's a problem for Mattermark because they're made of money, but the problem is that so much of what we consume, the content that we consume is ad supported. And I don't want to kill that. I think that's a great thing. You don't want paywalls all over the internet. That's not a solution.
Devindra: If you like the internet, ads suck but you'd probably want these things to exist.
Leo: And the other side of this is you're pushing ads underground and you're starting, you promote native content, you know ads that look like content which is far worse.
Alex: So I have a question. Do you run ad blockers yourself?
Leo: Yea, I run uBlock Origin.
Alex: Yea. I mean I ran ad blockers when I was at TechCrunch, I mean honestly. They didn't like that, but.
Leo: The reason I run it here is because I'm showing websites and so we don't want to have a lot of cluttered websites. But I have to admit—and I don't run it on mobile by the way which is really where you need ad blockers.
Alex: It is really, really bad. Browsing on your phone is almost unusable because you get pulled into application links and so forth. It's terrible.
Leo: I will submit though—
Greg: Even on Android, you're suddenly getting hack downloads happening automatically.
Alex: That is super, super gross.
Greg: So this new Android thing where they just, they install the app and they chunk the app download and people are bypassing that to do automatic—
Leo: That's not available yet until Nougat, or Nouga as your people would have it (laughing).
Greg: No, Australians say, Nouga.
Alex: It's about as bad as our pronunciation.
Alex: I mean I love everything else, just that pronunciation I do not like.
Leo: So that's not widely spread but it will be when Nouga becomes the standard. That's terrifying.
Greg: Yes. Think about that. Or think about what happens to—
Leo: So this arbitrary code can download and run?
Greg: Yea. There may be a—
Leo: There's got to be a way to fix this.
Greg: People are working on a way but that would be a real fear. Think about how much—for me it's just a speed thing. I go to websites and you know, the popovers, the popunders, the blockers. And I don't have time to wait for them to load. It's seconds, like 15, 20 seconds for a page for a major media property to load.
Alex: I'm curious if it's more data used for the ads to get on mobile, or is more data used for the actual content?
Greg: That's absolutely proven.
Leo: Probably for the ads.
Alex: Yea and that's very annoying to me.
Greg: So 2 seconds for the page to load and 15 seconds for the ads.
Greg: Yea. A lot of that is to do because it's a live transactional market so your data is actually extracted and goes into a live market. The impression is actually bought and sold, traded on multiple auction sites before the ads actually delivered to you.
Leo: Incidentally, speaking of Snapchat, what do you think of Instagram's Stories?
Devindra: It's definitely a blatant copy. They've admitted as much.
Leo: Well they don't care. Why shouldn't it be? That's fine.
Devindra: That's fine. I mean it's—
Leo: My problem is it's not navigable very well. So I thought it was a great idea, but now I have probably 20 stories up at the top there. I'm not going to go through all those stories. So I just ignore it. So their initial response apparently is quite good; however, I have a feeling it's going to tail off very soon.
Devindra: They're looking for more of a casual usage than Snapchat made popular, so.
Leo: I just think it's poorly implanted on the Instagram site.
Alex: I wonder if they're trying to do more to blunt or limit or mitigate Snapchat's growth curb than actually drive stories on Instagram itself. Why replicate and borrow?
Leo: Interesting. And maybe for brands, right. Hashtag brands.
Alex: Hashtag brands.
Leo: Hashtag brands, baby.
Alex: All those brands.
Leo: It's all about the brands, isn't it? I mean, wasn't it for Snapchat? The stories started really as a way for brands like Yahoo with Katie Couric to provide content on a Snapchat platform.
Alex: But whenever people uses Yahoo now as an example, I'm like, let me know when that ends. It's kind of like, yea.
Leo: Yea, it don't work so good. Yea. Oh. Yahoo. All right, let's see here. Allyn Malventano, you're next. Did you pick a story? Have you been thinking?
Allyn: I turned it into a shameless plug to wrap the show up. How about that?
Leo: I like it. I like it.
Allyn: All right. So it's the last one.
Leo: Ok, hang on, because we don't want to wrap the show up yet. It's Greg Ferro, has a chance to say a story as well.
Allyn: All right, yea.
Greg: So the one I like is that researchers cracked the Microsoft Secure Butte feature.
Leo: What's the Secure Butte feature?
Greg: Secure Boot. Sorry.
Alex: Who says butte?
Leo: Oh, yea, and did you love, did you love the mod scene presentation of that? I've got to show it. I've got to show you. So this is the golden key crack. What a big problem this is going to be for Microsoft.
Greg: Yea, so the idea was that computers that Windows sold would only be able to run Windows and the buyers would be locked with a security key that Microsoft had.
Leo: Now that is a little bit of a biased way to put it.
Leo: I think more—Microsoft's intent was not to keep you from running Linux but just to make sure that you didn't have a root kit, you didn't have a modified operating system, that you were in fact running a good version of Windows. You could say it was anti-competitive and I think some Linux users did.
Greg: Microsoft had choices about how they chose to implement this feature and they chose to use the anti-competitive feature.
Leo: Oh, all right. (laughing). Well, guess what? No good deed goes unpunished and no bad deed goes—this is the mod scene reveal of the bug.
Alex: That's awesome.
Leo: Isn't this hysterical. If you had an Amiga—turn up the music on this too. If you had an Amiga, you would recognize the style of music, the style of animation and even the Amiga typeface.
Alex: That is so awesome.
Leo: Now you can turn the music down.
Allyn: That's an old school intro for when you did games on the underground.
Leo: Yea, and the Amiga world, there were a lot of these circulating. They called it the mod scene. However, if you scroll down, you'll find an interesting hex block that apparently—
Allyn: Hmm, I wonder what that is.
Leo: It's the secure key. So what does this mean, Greg Ferro?
Greg: So first of all this means that people who bought Windows only computers can now unlock the secure key.
Leo: Woo hoo!
Greg: And put on whatever they like.
Leo: Is that hard to do because I'm curious. It sounds like it was a fairly long process.
Greg: It is I believe. I'm not an expert here because I don't actually care much about Windows. I just don't use it.
Leo: No, I care about putting Linux on Windows machines though. I like to do that.
Greg: So I suspect something will patch it up into a useful activity but the flip side of this of course is that malware authors or people of malicious intent now have the opportunity to take advantage of this and start accessing into your BIOS, right?
Leo: Oh, and that was the, at least Microsoft stated its intent for this, to keep people from mucking around with your boot code.
Greg: So fair enough. But what it also I think the more deeper ramification is this shows that cryptography algorithms where a magic key, the golden key, you mandate it, always fails. And it's not because the actual implementation, in this case you would have expected Microsoft to stuff up the implementation with really crappy programming because that's what they've done for the last 20 years so provably you'd expect them to repeat that pattern. But what they actually did is posted the code onto GitHub or something, like it was a public code depository.
Leo: He didn't even have to extract it? He just found it on GitHub?
Greg: Found it. I believe that's the story. I hope I got that right. So it's not even that someone did clever hacking to bypass. Apparently the code's pretty sound. It was just—for once the key's golden.
Leo: But that's the problem with these keys is that eventually they have to live somewhere in the clear to be usable. So even if it wasn't published on GitHub, there would be some way to extract that. That's what happened with DVD. That's what happened on Blu-Ray.
Greg: So somebody could just make sure that Theresa May in the UK knows about this and the Republican nut jobs here in the United States (laughing).
Leo: Well, Slipstream's beautiful mod scene code here actually has a paragraph about the FBI. "Are you reading this? If you are then this is a perfect real world example about why your idea of backdooring cryptosystems with a ‘secure golden key' is very bad! Smarter people than me have been telling this to you for so long, it seems you have your fingers in hour ears. You seriously don't understand still? Microsoft implemented a ‘secure golden key' system. And the golden keys got released from Microsoft's own stupidity. Now, what happens if you tell everyone to make a ‘secure golden key' system?"
Alex: I love how mean that is. Like I want to crack some things, throw some things. I want some other snide remarks for other agencies. He's not wrong, or she.
Allyn: That key packaged up with the right exploit means that your system with secure boot gets a root kit on it.
Leo: Well the other side of this almost all systems sold today you can disable secure booting in the startup, in what we used to call BIOS setup. I don't know what you call it now.
Leo: UFI setup. It's not even that. It's the setup code.
Leo: I don't even know what—I'm going to call it BIOS setup because everybody knows what that means. And you can just day, "It was only a handful of devices," like Surface RT that you couldn't bypass this. So good news now, you can put Linux on your Surface RT.
Alex: Because those sold so well 4 years ago.
Leo: (Laughing) If you're a hacker, hey, good news. You can now hack the 4 people that have—
Greg: You can turn them into a photo frame.
Leo: I have an RT machine it's cracked with a cracked screen.
Greg: For me it's the taking it to the British government who's trying to implement what they call the Snooper's Charter and you know, mandate that all encryption shall be crackable.
Leo: As far as I know, Theresa May does not listen to TWiT.
Alex: But Theresa, if you're on, we're big fans. Please don't ruin encryption.
Leo: Please. Is the Snooper's Charter back?
Greg: Yea, yea.
Leo: They're proposing it again? Because I remember, Cameron said, "No, we're not going to do this." But then he backpedaled a little bit and then of course Theresa May, I'm sure she—
Greg: Well she's the champion as that as the Home Secretary. And she's now—
Leo: She's the Prime Minister.
Greg: It passed the Lords House and it's in the Lords.
Leo: Oh, the Lords would never approve that.
Alex: Because the lords are so up to date on technology. They're not going to be as progressive.
Leo: (Laughing) What is this all about, I say?
Alex: Where's my ear trumpet?
Greg: With the opposition for the party in such disarray in the UK, it might actually get through with Brexit and all of that.
Leo: No kidding.
Greg: Yes. Some real concerns.
Leo: So what would be the impact of the Snooper's Charter getting through?
Greg: Basically that any encrypted communication—it would be illegal to create an uncrackable encryption, use uncrackable encryption.
Leo: As it is now in Russia by the way. Every, and I don't know what the upshot of this is. I don't even know how they enforce this, but Putin signed an executive order saying that any encryption used in Russia has to have, the company who created it or sold it or distributed it had to have a way of reading the clear text and passing it on to the Russians that have asked.
Greg: So we'll see. The question is there is a lot of laws that governments make that never get enforced.
Leo: Yea, I doubt very much that they've banned iPhones in the Soviet Union.
Greg: The flipside of this if I do something and the government decides that I'm—
Leo: They'd go after you.
Greg: I'm instantly a criminal for using safe cryptography and at that point they can take my computer because I've been breaching the law.
Leo: Well good news, iMessage is no longer safe so you're ok.
Alex: Wait. So because of Boris Johnson and the Leave Brigade, David Cameron's gone so we have somebody who's even more hawkish and digs technology even less so we're going to blame this all back to Boris.
Alex: Good. His hair is very annoying to me, so.
Greg: Yea, good old Boris.
Alex: Well, bad old Boris, but.
Leo: I think we could now go to Allyn Malvantano who will wrap this up and put a nice bow on it.
Allyn: It's not even a news story. It's kind of news.
Leo: What's the story?
Allyn: So me being a storage editor and reviewing SSDs and stuff all the time—it's probably the bottom link. I made my own test.
Leo: A test for what?
Allyn: To test SSDs and hard drives and stuff.
Allyn: I've been using this for a few months, kind of rolled it out.
Leo: It's a latency percentile test.
Allyn: Yea, it's called latency percentile, so—
Leo: What the hell is that?
Allyn: What you do is—so SSDs do a lot of things and most tests will give you the average performance throughout a given test. They'll run a test for a few minutes, and you get a specific number. What I noticed is there's a lot of SSDs out there that due to that poor firmware implementation we were talking about earlier, where they kind of get hung up on themselves and stuff around, they might hang for like a second or two every so often. But the problem is if you run a test over the course of a couple of minutes, then you have one or two of these little hangs in there.
Leo: It gets averaged out.
Allyn: It's averaged out. You don't know this.
Leo: Would I notice the hang as a user?
Leo: That's the pause as I'm using the computer?
Allyn: Yea when you're trying to do something, you're trying to open up an app or something and it just kind of sits there for a second or too because it's busy waiting for some of the IOs to complete, right. SO the key is a few of these requests will just take a little bit longer than all of the rest of them, right. The only way to track that and find that is you have to look at every IO. Every single one.
Allyn: You have to log all of them. So I test SSDs that do upwards of, like the fastest ones do like a million IOs per second.
Leo: A second.
Allyn: A second.
Leo: You can't log that.
Allyn: I do.
Leo: How would you log that?
Allyn: That's my secret sauce. I figured out how to do it.
Leo: In memory and RAM (laughing).
Allyn: Basically, yea. But so what I do is—
Leo: Wow. So is that what that spike is for instance? Is that a pause right there?
Allyn: So what you're looking at there is the source data for—I generate the percentile results. And those are a bunch of different items and if you scroll down a little bit. So I've kind of turned that into what's called a percentile and it's showing, ok, where is the point where all of the IOs fall on a SSD? How long do all of them take. And the things on the left there are NVME SSDs. And then—
Leo: Look at that.
Allyn: A little bit closer to them, a little bit to the right is SATA SSDs and there's a really old SATA SSD that's right down the center, that gray one, right. And then the 3 things on the right—
Leo: These are spinning drives.
Allyn: Three different RPM hard drives, right?
Leo: Isn't that funny. So we were buying these 10,000 RMP hard drives to get these micro improvements when if we just bought SSDs, even crappy SSDs.
Allyn: And that's a long scale so every time you get to a major increment—
Leo: Oh, no.
Allyn: Yea, you're times 10 every time you go a major increment.
Leo: So those 10,000 RPM drives were 100 times faster. However, we're talking thousands.
Allyn: Yea, way faster. And when the XPoint stuff comes out it's going to be even closer to the left edge of that access.
Leo: Oh, wow. Look at that.
Allyn: So where it comes into play is testing SSDs and finding the ones that do, that kind of hang for periods of time.
Leo: Not all of them do that.
Allyn: Not all of them do that. That's just the example I showed to explain the technology but if you go to the next page, yea if you go to the next page in the article, those are like the top.
Leo: I need to go to the end.
Allyn: Yea. And you go to the 2nd to the bottom picture. Where are you going?
Leo: Well there's a lot of comments here. I shouldn't have gone to the end. Should I go to the top?
Allyn: Yea go to the top.
Leo: I don't understand how these slideshow works. Occlusion?
Allyn: Next page.
Leo: Latency weighted percentiles. Yea, here we go.
Allyn: So that's where I'm comparing the drives. And then if you scroll down to like 2nd to the last picture, those are reads. Reads tend to be pretty much equivalent across different models but if you scroll down a little, you start getting rights.
Leo: You see a difference. Oh, look at this.
Allyn: Ok. So now you see that there's a SSD there that's doing something that the other ones—
Leo: What's that little one doing? That little stair step thing?
Allyn: That has a significant number of IOs that are taking a very long time.
Allyn: And they're actually getting far enough over in that chart to where the hard drives were operating before.
Leo: So don't buy the Plextor.
Allyn: The Plextor M6e was the one that didn't do very well. And if you scroll down a little more—
Leo: This is great that you wrote this because I don't think this would be obvious at all in any other test.
Allyn: Those kinds of hiccups are things that totally get lost in the wash if you just di the average performance for those things. Yea, and that's as you put more load on the drives, even some of the better ones start to become a little bit worse.
Leo: Yea, that's interesting. Yea. Look at that.
Allyn: It looks like a New York Subway map once you start getting all kinds of crazy.
Leo: (Laughing) Wow this is really interesting. So, what do you like?
Allyn: What do I like?
Leo: The OCZ? No.
Allyn: The 950 Pro's still good.
Leo: The 950 Pro. That's the Samsung.
Allyn: Even the Samsung 850 EVO drives. Anything over 500GB, those guys are amazing. They saturate SATA—like if you don't have a need for MVME—
Leo: That's what I'm using with my ZSF setup for those.
Allyn: The 850 EVOs?
Leo: Yea. Yea. The grade Z mode. It works pretty nice.
Allyn: I don't even recommend the 850 Pros because the EVOs have, they treat some of their flash as SLC which is really superfast. So in some respects, the 850 EVOs outdo the Pro model believe it or not.
Leo: Wow. You know I noticed that it was a little hard to figure out whether to get the EVOs or the PROs because of that. It wasn't obvious if one was better than the other for some things.
Allyn: For SATA, EVO is the better way to go. And it's almost half the cost.
Leo: Yea, it's a lot cheaper.
Devindra: So what's the advantage of the M.2 drives because I've been seeing those come about and I'm a guy who follows PC hardware. These things appeared a couple of years ago. I'm like, "What, where does this go?"
Leo: Well, you're using MVME, right?
Allyn: Just because you say M.2 does not equal MVME.
Leo: The ones I've been using are all MVME but not necessarily. They could be SATA.
Allyn: So, yes, there is such a thing as M.2 SATA. And that's where it really started because people were looking for, you know, really slim form factor, notebooks that couldn't even fit a 2.5" form factor drive. They wanted something more compact. That was the first iteration. It was supposed to be called Next Generation Form Factor. That's what the NGIF thing which turned into M.2. So initially it was just make things smaller. And then they included, because they were smart, forward thinking, they included PCIE into the spec for the possibility to connect it. And then Intel pushed MVME as a—basically it's—
Leo: So you want all 3. You want M.2, PCIE and MVME.
Allyn: Yea, all three of those things together are all like really small compact SSE.
Leo: And by the way, all I buy now in desktop and laptop because it's notably much—
Allyn: It's very, very fast.
Devindra: It's just, it's hilarious too because it's a tiny little chip. And you plug it in. Replacing even the 2.5" SSD. It's like what is—and then if you—
Leo: This is an MVME as well as a—
Allyn: You just have much more bandwidth. You have the equivalent of—well, you have a rate of 4 850 EVOs at home, right?
Allyn: That is roughly the equivalent of a single M.2 MVME SSD.
Leo: Wow. Wow. And that machine boots off of a M.2, 512GB M.2 and then it uses for data the—
Allyn: So yea, if you're moving stuff back and forth between one and the other, you're actually almost equivalent in speed between both sides which is good, right? If you need to back up your—
Leo: (Laughing) no, that's what—yea, I'm using it for snapshots and things like that, yea. That's exactly what I'm using it for.
Allyn: So you're in the ideal situation. Your copies are going as fast as they can go.
Leo: Yea. Yea. Nice.
Devindra: Good to know.
Leo: You know how I learned that? From you building the ultimate reality gaming machine.
Allyn: That's true. Which is at your house now, right?
Leo: Which is now at my house where it has a very nice home. Very nice home. Hey, this was a lot of fun. We went long because we had such great stuff. And I'm really glad you were all here. Let's all go to dinner. Alex Wilhelm, he is at Mattermark.com.
Alex: Yes, sir.
Leo: And he's not working at Jet despite the uniform.
Alex: Actually I have Jet socks too.
Leo: That's nice.
Alex: They send out weird swag when they launched. I don't know. It's almost like they spent too much money before they got up. Who would've thought?
Leo: You know the rule is always make the t-shirt then the product. So did you have anything you want to plug at Mattermark? Can people read it without subscribing?
Alex: Yea, they can. Just mattermark.com/blog and click on anything from editorial. That's stuff I'm either writing or co-publishing, so.
Alex: That's where I'm working. I'm on Twitter @devindra if you want to follow along.
Leo: (Laughing). He's @alex. @alex. He loves it. He's the only Alex on Twitter. Nobody else.
Allyn: How did you get that?
Alex: I bought it off a guy in Mexico in 2008.
Allyn: Oh, seriously? Wow.
Leo: Yea, because Alex was a 3 instead of—
Alex: No, different Alex. But he kept that so I was @alexwilhelm and then the guy in Mexico wanted to sell it and I had $60 bucks in my PayPal account, which was all the money I had at the time to my name.
Devindra: Smartest investment you've ever made.
Alex: This is back when Friendfeed was a Twitter competitor. So it was a really different era of social. And so I sent him my $60 dollars. And he changed his name and I quickly changed mine to @alex and it worked.
Leo: So you're real name is Timmy?
Alex: My real name's Bill actually. But the pronunciation comes out Alex in common English. So don't be racist is what I'm saying.
Leo: No, no. Devindra Hardawar. He's Senior Editor at Engadget, the great gadget blog and he's out here for IDF.
Devindra: I was out here for IDF.
Leo: I didn't get to ask you about the Intel Developer's Forum.
Devindra: Yea, it's maybe we'll hear more about their next chips that we've been waiting so long for, right?
Leo: Kaby Lake. What was the other one? Kaby Lake and—
Allyn: That's one of them.
Greg: A rock or something.
Leo: Nobody even knows. Nobody even knows.
Greg: Well the names are kind of like—
Leo: They all blur together. Huh?
Allyn: It hasn't happened yet.
Devindra: It happens this week.
Leo: Oh, IDF's coming up, that's right. Well, enjoy IDF. We'll talk about it next week. Thank you for being here. It's really nice to have you.
Leo: Same here, it's great to have Allyn Malventano. Always a pleasure when you're in the studio.
Allyn: Happy to be here.
Leo: Please give the best to your friends at the NSA and tell them I didn't mean—
Allyn: I'm retired.
Devindra: They know, they know, Leo.
Leo: I didn't mean all those nasty things.
Allyn: They make you sign so much stuff when you leave, you're not talking to any of them.
Leo: You're done. You're gone. You're a non-person.
Greg: We'll need to get together later on and talk nuclear. I used to do nuclear power (laughing).
Leo: He is number 9. Don't ask who number 2 is.
Allyn: I used to run reactor plants on submarines, so we can talk.
Leo: Oh, there you go. And for some reason I get the feeling that there's nothing Greg Ferro can't talk about, so.
Greg: I just make stuff up.
Leo: Master of—he just wings it. You can find it at packetpushers.net. This I am sad to say is the last TWiT we'll be doing in this studio. I'm glad you all, a great studio audience, could be here for this. We will next week be doing the very first show. There's some question about that since apparently there's an electrical issue. We may be in the dark but—it might be outdoors.
Alex: Oh, that would be awesome.
Allyn: Get a generator.
Alex: A nice little hum in the background the entire time.
Leo: I will make it happen somehow. I'll find a way to make that studio work next week. So do tune in early because we'll be taking the trolley car over to the new studio at about, I'd say about 2:30, between 2:00 and 2:30 Pacific, 5:00, 5:30 Eastern, 2130 UTC. Join us for TWiT next week and every week. If you can't be here live you can always get after the fact on demand audio or video from our website twit.tv or wherever you subscribe to podcasts. Please subscribe. We want to make sure you don't miss an episode. Thank you for being here for the last time in the TWiT Brickhouse, I say, another TWiT is in the can! Bye-bye.