The switch to USB-C

▼ A week ago, Apple unveiled the new MacBook. The crazy thing about this new computer is that it only has two ports: an audio port and a USB type C port for everything else. The audio port is like the one in the MacBook Pros: it looks like a regular 3.5 mm headphone port, but it also supports iPhone-style headphones with a microphone and clicker, and with the right adapter, it does optical digital audio out.

The USB-C connector is very interesting. It's backward compatible with old USB standards at 1.5, 12 and 480 Mbps with a simple adapter cable. It also supports USB 3 and 3.1 at 5 and 10 Gbps, and should be ready for even higher speeds in the future. Like Apple's lightning connector, you can plug it in in both orientations. USB-C can deliver up to 5 amps at 12 or 20 V for a maximum of 100 Watts—in either direction. Last but not least, the USB-C port allows many non-USB protocols to be transmitted over a USB-C cable. So with the right cable, a USB-C port can be used to connect a monitor over HDMI or DisplayPort. This uses wires normally used for 5 Gbps USB connectivity. Depending on how many wires are used for the alternate protocol, the USB speed may have to fall back to 480 Mbps.

The new MacBook is a lot like the first MacBook Air: that one also only had a USB port and an audio port, in addition to a power connector. It was also unbelievably small and light for the time, but also underpowered and very expensive. The new MacBook has a mobile CPU that only draw 5 W, so it doesn't need a fan. This probably means that the machine can be fast only in short bursts, as it will have to throttle back when it heats up. We'll see when the benchmarks come in. In most respects, the new MacBook compares favorably to the existing 11" MacBook Air: same battery life at 9 hours, and the slightly larger screen is HiDPI at 2304x1440 pixels, compared to 1356x768 for the 11" MBA. However, that's only 1152x720 "points", so less stuff will fit on the screen. I'm already having problems with that on my 1280x800 point MacBook Pro. I can't imagine having to shoehorn my digital life into a 19% smaller number of pixels.

But back to that single USB-C port. Like the iPhone and iPad, the new MacBook uses the same port for charging as well as expansion. So if you want to connect your iPhone or a USB flashdrive to your MacBook, you'd have to disconnect the power supply. Until a few years ago, that would have been very problematic: laptops had relatively poor battery life, so they were hooked up to power at every opportunity. But today's laptops have battery life similar to that of an iPad. It's not entirely inconceivable to use them on the battery during the day and charge them overnight. Pulling out the power to sync an iPhone or copy some files to an USB drive is not a big deal. Of course for the time being, you'll have to use a USB type C to type A adapter, like the one Apple sells for $19.

And if you want or need more long-lived USB connectivity, Apple has a two additional adapters that allow you to connect regular type A USB cables/devices to the new MacBook while the computer is connected to power. In addition to that, one adapter has VGA out, the other HDMI out, allowing you to connect an external screen at resolutions up to 1920x1080. (These adapters aren't cheap at $79.) On the tech specs page, Apple also mentions "Native DisplayPort 1.2 video output", but there's no cable or adapter for that yet.

All of this would be a lot easier if the new MacBook had two USB-C ports. Think about how great it would be if you could plug in power at either side of the computer! And then you still have a free port to connect USB devices and/or an external screen. Because USB-C is an industry standard, it will be possible to buy third party power supplies, which will probably be a lot cheaper than Apple's. And having a single power supply that can charge pretty much everything will be great. However, I'm slightly worried about the loss of magsafe. I once tripped over my PowerBook power cable, and the computer went flying through the room. Hopefully, the USB-C power cable is loose enough that it comes out in such a situation.

The new MacBook's keyboard extends all the way to the sides of the computer, and the computer is so thin that it's not possible to have connectors under the keyboard. This leaves very little room for connectors at the far end of the computer, in line with the hinge, with the USB-C connector on the left and the audio connector on the right. So adding a second USB-C connector probably means the audio connector has to go. But that's easily solved with a small USB-to-audio adapter. Apple could even make a magsafe+audio-to-USB-C adapter.

The MacBook Air quickly gained a second USB port. It's certainly possible that the new MacBook will also gain a second USB port by the time the price drops and it replaces the MacBook Air as the base model in Apple's laptop line-up. (Which I'm pretty sure is the goal, considering the MacBook <nothing> name.)

So how is the switch to USB-C going to work for the rest of Apple's line-up?

Will the next iPhones and iPads have a USB-C port rather than a lightning port? Could be. But it seems a bit too soon for that, especially because this is an "S" year for the iPhone. So maybe next year. It doesn't make much sense for Apple to include a USB-C to lightning cable with i-devices at this point. But then again, how long are they going to include USB-A to lightning cables? Perhaps you'll be able to choose? I do expect them to start selling USB-C to lightning cables separately later this year, especially if they don't plan on switching the i-devices to USB-C the year after.

I assume Apple is going to see how well USB-C works on the new MacBook before putting the port in the MacBook Pros. The MBPs aren't going to shrink so much that the USB type A port doesn't fit anymore anytime soon. And the thunderbolt ports support 20 Gbps today while USB still has to implement the 10 Gbps spec. However, once DisplayPort 1.3 can run over USB-C, it will be possible to run a 5k display over a USB-C port, so Apple will be able to build a 5k Cinema Display, which can then also power the computer over USB-C.

In an ideal world, a new version of the MacBook Pro would have two USB-C ports with DisplayPort 1.3 and USB power capability, while still maintaining the existing magsafe and thunderbolt ports. However, it seems unlikely that we'll be that lucky.

But one thing is for sure: we're going to need a whole lot of new cables. Full USB-C needs 24 wires, and the cables may be active cables with a chip in them. Which means they won't be cheap. But there's probably going to be simpler cables that only do 480 Mbps data or even no data, but only power using a USB-C connector.

Permalink - posted 2015-03-16

Canon PIXMA printer vs Apple's photo lab: it's pretty much a tie

▼ A few years ago, I ordered some prints through iPhoto. I've also used other services to order prints of digital photos in the past, and those always look just as good as "real" photos, as long as the original file is of sufficient quality. But what about my new Canon PIXMA MG7550? How do photos printed using it compare to photos printed by a lab? See for yourself:

One of these is a print from the lab that Apple uses, the other was printed using the MG7550. Can you tell which is which?

The top one is the one from the lab. If you look carefully, you'll see that the blacks aren't as dark and the colors aren't quite as saturated. However, the bottom one has more of the image cropped off. The printer has no problem printing borderless, but it only has a limited number of size presets, so for 15x10 cm photo paper you have to use the 6x4 inch setting, but that's actually 15.24x10.16 cm, so the paper is 1.5% smaller than what the printer thinks.

The image above is a scan at 600 DPI (24 pixels per mm) of the two photos. If we zoom in to 200% on the two photos, it becomes clear that the lab photo is sharper and has smoother colors than the home-printed photo.

Printed by Apple's lab

Printed on the Canon PIXMA MG7550

The original JPEG from the camera

However, this is a very small detail enlarged about 12x. I can't tell sharpness/smoothness difference with the naked eye. The lab and the printer both have strong and weak points, so I'm declaring it a tie. However, the strong points of the printer are much easier to observe.

Also note that I used HP Premium Plus photo paper, but I told the printer it's Canon Photo Paper Plus Glossy Ⅱ. The printer comes with three sheets of Glossy Ⅱ paper, and prints on that are actually even more contrasty. When selecting the "other photo paper" setting with the HP photo paper, the result is markedly less contrasty and yellower. You'll get the most accurate results by creating a color profile for the printer/ink/paper combination you intend to use.

So the results are pretty similar. But the prices aren't. Photo paper gets expensive fast: the Canon paper costs 8 euros for 20 sheets at a local store, that's 40 cents per sheet! Even online you pay about 18 cents per sheet. HP is a bit cheaper and there are also no-name variants that get below 10 cents per sheet. Don't forget you also need to buy ink, which will also be several cents per photo.

Ordering photos from a lab is much, much cheaper. From Apple through iPhoto, you pay 12 cents per photo at 15x10 cm. However, there's an additional € 4.83 per order shipping costs. So you need to order at least three dozen photos to reach a per-photo cost of 25 cents. This is similar for other services.

So if you can avoid paying inflated prices for your photo paper, printing at home makes sense if you value being able to print a few photos here and there rather than ordering relatively large numbers at once. Also, there's the instant gratification angle. Having to wait several days or even a week to see your prints sucks, especially compared to printing wirelessly from your iPhone in 20 seconds.

Permalink - posted 2015-03-06

More Apple Watch predictions

▼ Unless something really unexpected happens, Apple will further unveil the watch on Monday. I also think it's safe to assume the event will be streamed. If you're in Europe, remember that the time difference is an hour less than usual because the US will "spring forward" to daylight saving time this weekend while we'll be on winter time a few more weeks in Europe. So the event starts at 18:00 CET.

Here are some musings and predictions.

First of all, I'm pretty sure there's another shoe that has yet to drop. There will be something significant that hasn't been revealed yet. Another sensor? Maybe it'll support GPS after all? Less iPhone reliance than we currently assume? An additional cool use case? I don't know. But I'll be surprised if there isn't something.

I think pairing with an iPhone 6 or 6+ will happen through NFC: start the watch app, select "set up an watch", touch the two devices, and you're done. Pairing with an older iPhone will probably be more like regular bluetooth pairing.

I wouldn't be surprised if a software update will be needed out of the box. Apple is probably working on the software until the last minute while factories have been cranking out watches for some time. If that's the case, there's really no need for a staggered international release, so the watch could be available in a good number of countries from day one. (I can only hope.)

Battery life is a big unknown. The watch can do a lot of things that require a fair bit of power, so its tiny battery could easily run out during the day. On the other hand, most of the time the watch will be sleeping, doing nothing more than monitor the acceleration sensors and the bluetooth link, neither of which require much power. And I'm sure Apple realizes that having to charge the watch halfway through the day would be extremely problematic. With a phone, topping up the battery when you're around a power outlet isn't too big a deal. But with a watch, you'll have to take it off, so you can't really use it while it charges. The watch absolutely needs to be able to get through the day on a charge for nearly all users.

We know charging will happen through an inductive connection. What we don't know is what's on the other end of the charging cable. Maybe there won't even be a cable, and you'll have to use your existing lightning cable to charge the watch. That would suck big time, because then you can't charge your iPhone and the watch together overnight. Hopefully Apple won't be that cheap. The opposite would be that the watch comes with its own charger. That seems somewhat unlikely, too.

I remember people mentioning that the watch's inductive charging system uses a lightning cable, but I don't see that on Apple's website and it doesn't seem very likely. I think the inductor will have a cable attached.

What makes the most sense to me is that the watch's charging cable is plugged into an existing iPhone charger and allows the iPhone's lightning cable to be daisy chained, so the phone and the watch can be charged together from the same charger.

Then there's the price. People have been speculating about truly stratospheric prices. However, if Apple is going to sell watches for $20000 or more, this is likely going to alienate their regular customers to some degree. Also, even though the profit margin is enormous, the infrastructure required to sell such a super luxury product would be significant, and could very well be not worth it. Just like with the iPad, Apple is probably going to want to disrupt the high end watch market with relatively low prices for the gold watch. I stand by my earlier prediction:

I'm thinking just like Apple is dragging up the low end, they'll probably drag down the high end. $3500 sounds like a nice, round number that still quite a few people can afford. Maybe they'll go a bit higher, but I doubt they'll break the $10k barrier.

I mentioned a price range of 700 to 1000 dollars for the steel version. I'm hoping maybe it'll be 500 and then I'll consider that version, if it's not too heavy. However, $80 for bands seems on the low side now, looking at Apple's website it's obvious that they'll want to use the bands to be a differentiator, so those will get expensive fast. They may not even sell them separately, but if they do, expect to pay a lot except maybe for an extra sports band. I also wonder if we'll see third party bands, or if Apple has a way to keep those off the market.

I wonder if the 42 mm model will cost more than the 38 mm model. In the past, bigger devices have invariably been more expensive. I wouldn't put charging $350 for the 38 mm sports model and $400 for the 42 mm one past Apple. On the other hand, this is uncharted territory, so the old rules are out the window. It makes sense to me to sell the two sizes at the same price, except maybe for the gold models.

There's also the storage tiers that Apple has traditionally used to differentiate pricing. I doubt that they'll want to further confuse things; my guess is that all the watches will have 8 GB storage.

Permalink - posted 2015-03-05

The Canon PIXMA MG7550 and the state of printers in 2015

▼ Today, I bought a new printer: a Canon PIXMA MG7550.

First, let me tell you about my old printer. This is a HP Deskjet 5652, which is from around 2003. Back then, it was a really nice printer: it prints borderless photos and it has a duplex module for double sided printing. Hardware-wise, it's still in perfect working order after a decade. You can even still get the cartridges. The trouble is that the latest MacOS version that has drivers for the 5652 is 10.5 Leopard. You can still use the printer with generic Deskjet drivers under 10.6 and later, but then you only get basic printing using the standard settings and nothing else. So when I want to print photos, I have to boot into 10.5. That gets old fast.

Also, the Deskjet 5652 connects using USB or a parallel (Centronics) port. Not a big problem, as the Apple Airport Extreme and Time Capsule Wi-Fi base stations have a parallel port for precisely this purpose. However, you then still need the right drivers and you can't print from an iOS device using AirPrint.

Due to my limited printing needs I accepted these limitations. However, the other day I noticed that some text I was printing kept coming out looking smeary unless I printed in black and white. Also, the color cartridge had run out of cyan, so I was looking at a € 45 or so investment into a new cartridge. So I did some research, and found Canon's PIXMA line.

There are very many PIXMAs. I think they're mostly lower-end photo printers. You can get one for well under € 100, but I decided to get a higher-end model: the MG7550. This one has 10/100 Mbps Ethernet as well as 802.11b/g/n Wi-Fi and USB 2 (no USB cable included, though). It also has a built-in scanner and it can print on printable CDs/DVDs/blu-rays. I got one at the local electronics store for € 134. It looks like it costs a bit more online, probably in no small part because the thing is huge and weighs 10 kilos, so that free shipping must eat into the online retailer's margins quite a bit.

I hooked the printer up using an Ethernet cable and installed the six ink "tanks". Apparently, unlike the inkjet printers I've used previously, the MG7550 has a fixed print head and the ink cartridges just provide ink. There's the requisite yellow, cyan and magenta, as well as gray, black and pigment black. The regular black is dye-based and mainly used for photos. The pigment-based black is mainly used for text. This is all selected automatically. I like the fact that there's a separate cartridge for each color, so you can replace them individually as they run out. (Yes, the cartridges are sold individually.) Each tank has a built-in LED that comes on when it's inserted correctly.

Canon wants to you to install software from a CD (Windows), their website (Mac) or the app store (iOS) to set up the printer. Not sure what that does, because after answering a few questions using the printer's touch screen and the color alignment procedure, the printer is ready to be used. The Mac discovers the printer using Bonjour. Because the printer supports the AirPrint protocol, there's no need to install any drivers—that's only necessary for the scanner part. The Mac automatically downloads those drivers from Apple.

Printing from the iPhone is even easier: select Share and then Print, the printer shows up... that's it. (It doesn't matter that the iPhone is connected over Wi-Fi and the printer over Ethernet.) The printer has two paper trays: a big one that holds A4 paper, and a second, smaller one that holds 15x10 cm photo paper. When printing from the iPhone, you simply select 15x10 cm / 6x4 inch and the printer will use the appropriate paper. Printing from the iPhone quickly became my preferred option, as printing from the computer involves a lot more settings, although printing from iPhoto is relatively simple. You can also print from an SD card, which is nice, but not quite as easy as using the iPhone.

The touch screen is really cool, and removes the need for inscrutable messages. When you load paper, the printer asks you to select the size and type, and then warns you when you try to print using a different size or type of paper. Printing text is pretty fast. Photos take a bit longer, but it's not too bad. When the printer starts printing, a small tray flips down to catch the printouts. The prints are really sharp and photos are very colorful, but the black could stand to be blacker, both for text on plain paper and for photos. I'll have to try Canon's photo paper to see if the results are better than with my HP and no-name photo papers.

Most people probably don't really need a printer anymore these days. However, if you do any printing at all, you really should get a modern printer: they're super cheap and produce very nice output, while AirPrint and Google Cloud Print have finally solved the driver issue. And with 15x10 cm photo paper loaded I can wirelessly print photos from my iPhone when the mood strikes.

IPv6 and firewalling

When perusing the settings, I noticed that the printer supports IPv6. I posted this photo to Twitter:

Only after I posted the photo, I realized that the printer's IPv6 address is now known to the world, so presumably, everyone will be able to print to my printer. After digging through the menus, it turns out there's no way to set an IPv6 address manually, and the printer doesn't use privacy addresses, either. So no security through obscurity for me. There's also no security settings to disallow remote printing. So I basically have three options:

• Do nothing and see what happens
• Set up an IPv6 firewall between my LAN and the IPv6 internet
• Disable IPv6 on the printer

None of these options are particularly great. I wish Canon had given all of this some more thought.

Update: this is what you get when you challenge the internet to print on your printer over IPv6:

@iljitsch @jaredmauch @danyork @appelboor Cool, that’s me ;-)

— Niels (@nlsp) March 3, 2015

Permalink - posted 2015-03-03

A battery breakthrough would explain the Apple car

▼ The latest Apple rumor is that they're working on a car. A rumor that is easily dismissed:

The fantastic Apple Car is a fantasy.

But is it? What would it take for Apple to start making cars?

Cars are a huge market and they're very expensive. That's exactly the type of market Apple needs to enter if it wants to continue growing its already unprecedented profits. However, cars aren't very profitable for the companies that make them today. So Apple needs to bring something unique to the table to allow it to reach the margins it's grown accustomed to.

There are already very nice high-end cars, and there's just not that many people who can afford them; selling a better car for a higher price isn't going to give Apple the numbers that it needs. What Apple needs to do is make a better car cheaper, so it can be sold for a regular price but at an Apple-like margin.

The insides and outsides of a car are determined by its function and the fact that it must push through the air at a relatively high speed. There's not much that Apple can do here. Maybe Apple has come up with some new user (driver?) interface paradigm that will revolutionize driving, but somehow I doubt that.

Maybe an Apple car is going to be self-driving? Not impossible. But Google has been working on that for some time, and they use existing cars. Also, Google's driverless cars are described as "trains running on invisible tracks": they're not very autonomous, but rely on extremely detailed maps. Is Apple confident enough in its mapping abilities to think they can pull this off?

But maybe they just build a better car. Apple isn't going to improve the motors used in electric cars, simply because those are already pretty much perfect: they weigh a fraction of what an internal combustion engine weighs, run at 90% efficiency and they last pretty much forever with little or no maintenance. Most of the other electronic and mechanical parts are small potatoes.

However, there is one thing in an electric car that could be improved: the battery. A better battery that stores more energy for the same size/weight would be a big deal. Or that can be charged faster. Or lasts longer before it has to be replaced. Or is simply cheaper.

We read about research breakthroughs in battery technology all the time, but we rarely see those translate into real-world improvements. Obviously, Apple has been interested in batteries for a long time. So what if Apple has developed a significantly better battery? That's sure going to help the Apple Watch, and we could all use a bit more battery life in our iPhones.

But if you're looking to monetize a battery technology breakthrough, look no further than the electric car.

Permalink - posted 2015-02-16