Wesley's Tool-Box

SK Telecom, the biggest mobile carrier in Korea, is the parent company of SK Planet, which owns Korea's 2nd largest online shopping mall, 11st (11번가, 11th Street). So it was natural to see that SKT would come up with a clone of Amazon Dash, which simplifies online ordering with the push of a physical button. It's called "Smart Button Kkuk(스마트 버튼 꾹)", and became available in September 19, 2016, about 18 months after Amazon Dash's announcement.

Initially affiliated with about 60 items in 11st and given away for free with first 50,000 orders of the website's "Now Delivery(NOW배송)" service (this itself is 11st's version of "Fulfilled by Amazon"), it's now able to order from more than 140 items and shipped for KRW100 (US$0.09) when ordering certain specially marked items through Now Delivery. As of this writing, the stock of these buttons have apparently run out, but I expect it to be replenished soon.


As with other SKT Smart Home devices, the button comes in a brown box with standardized instruction manual included. Looking at the device, the red button with 11st's logo on the left side is what you use to both pair and eventually order items online. The body is elongated to the right to house the battery. Like Amazon Dash, a removable hook is provided so that it can be hung.

Lots of companies are hopping on the wave of Internet of Things (IoT) these days, each trying to build an ecosystem of their own. Major Korean mobile carriers (SK Telecom(SKT), KT, and LGU+) are no different, each setting up proprietary efforts. I used to largely ignore them because they required monthly fee, ranging from a few dollars per device to tens of dollars for "unlimited" (discounted with multi-year contract). With many non-carrier solutions usable without fees, I felt this to be undesirable.

However, SKT must have felt that this arrangement was hurting their chances for wide acceptance, because it began offering some of their affiliated products free of monthly fees, charging only a nominal (KRW 5,500, US$4.80) one-time fee included in the sales price since last October. I decided to take the bait and started integrating their home IoT products, sold under the "SKT Smart Home" brand, into my house. The very first one was a smart plug. The one you see here is the one made by Witty.


Over time, I installed several of these plugs around the house and some of them came from another company, Dawon DNS. Because SKT was selling them, lots of things were made to be nearly identical, namely the packaging and the functionality. The frugal-looking brown paper boxes contain the plug and the instruction manuals, all of which look almost exactly the same except for the diagram of the plug. Both companies' plugs are controlled by the one and the same official app, and works largely like as if they are the same product.

Even the price is similar, at around $22.50 to $26, which makes them one of the most inexpensive options. This includes the aforementioned fee, so the pricing is rather aggressive. The deal is made even sweeter by the fact that the plugs do not need any gateways or hubs to operate as some other solutions do (namely that of other carriers'). It will interface directly via Wi-Fi, so if you already have a wireless router at home there's no additional cost involved.

I was quite surprised to see my Apple Watch Series 2 having more than 60% of battery on the first day of use. To see that this was not a fluke, I kept checking for a few more days and realized that it usually had 50% or more left after 24 hours with light use. In such cases, I was able to go without recharging for two full days. This is quite a bit longer than the Apple Watch Gen 1 even compared to its early days. So to make this clear, I did a comparative battery discharging test to produce the graph below. The devices had watchOS 3.0 installed at the time.


Apple Watch Gen 1 lasted 24 hours 45 minutes, while Series 2 worked for 38 hour 50 minutes, about 57% longer. For Series 2, it still effectively meant two days' use - have it fully charged on the morning of day 1 and it will last until late evening on day 2. You can also see that if there was less activity, 48-hour use would have been possible as well. What's more interesting is how much battery is consumed for certain activities.


During regular office work, the drain rate is similar for both watches. But once they're subject to more demanding tasks like logging an exercise or actively running apps, Gen 1 tends to drain about twice as fast. The idle state shown by the sleeping time is also less efficient compared to Series 2.

Early parts leak showed that Apple Watch Series 2 42mm models have about 36% larger battery compared to Gen 1 (334mAh, from 246mAh). So the difference isn't just coming from a larger battery, but an even more energy-efficient system overall. Considering that Series 2 has a CPU twice as fast and a screen twice as bright, this is quite a feat.

Now, it's been suggested that Apple put a larger battery on Series 2 because it includes a GPS module. This would enable path logging without a paired iPhone at an expense of a faster battery drain. So let's see how much difference it makes.

To compare the battery life of the Gen 1 and Series 2 Apple Watches in a fair manner, I wore both devices on the same arm for as long as the device's battery lasted. In doing so, I was also able to see if there were any differences in the logged activity data. I'll be talking about this first.

I should note that this test required carrying two iPhones, one for each watch. While Apple did introduce 'auto switch' feature in iOS 9.3 and watchOS 2.2 to let a user use multiple Apple Watches, only one watch can be active at a time. So to log the activities independently and simultaneously, I paired Series 2 to iPhone 7 Plus and Gen 1 to iPhone 7. The phones were in my pants so that they would always be close to the watches. Both had been through more than 20 minutes of calibration walks before the testing as well. Let's see what the final results of this walking session.


The calorie and heart rate measurements were nearly the same, meaning that both Apple Watches saw nearly the same amount of activity. This is assuring as the measurement back-end for the workout session is consistent across generations.

One of the reasons why the 1st generation (Gen 1) Apple Watch felt limited was due to its noticeably slow performance. App performance, especially the 3rd party ones, was sluggish enough that I eventually gave up making much use of them. The watchOS 3 update has implemented a lot of optimizations that alleviate the problem, but it couldn't hide the fact that the CPU simply was underpowered. Apple Watch Series 2 aimed to fix this by putting a CPU twice as powerful. Let's see if it worked out. First up is the boot speed. All tests were done with watchOS 3.0 installed.


Compared to iPhones or modern computers, Apple Watches are notoriously slow to boot up. Gen 1 still takes more than two minutes and while Series 2 is faster, it isn't significantly so. Fortunately, you won't need to reboot it often, so it's a minor annoyance at best. If you want to see the boot speed in real time, here's the boring video.