Log in

No account? Create an account
Joshua's weblog
September 20th, 2015
09:51 pm


Previous Entry Share Flag Next Entry
Two weeks with a Pebble Time

It's been over a year since I started wearing my first-generation Pebble Steel. I expected that I was going to find it an annoying gadget that didn't really do anything positive for me, but I was pleasantly surprised, and somewhat taken by it, really; I wrote a few programs for it, and up until a week ago, I wore it around with me pretty much everywhere. Last Friday, however, my Kickstarter reward arrived: a shiny new Pebble Time Steel in gunmetal black! The Pebble Time is Pebble's second-generation smartwatch, with a host of changes: color, a microphone, a new industrial design, and a new user interface layer. I've been wearing it for just over a week, then, and so it seems apropos to write another review, in the same vein of that which I wrote in March of 2014, of the first-generation device.

So let's get started, then. I promised you just a short sentence ago that I'd write a review of how I felt about the Pebble Time Steel (which I'll just call the 'Time' from now on), but to put it all into context, I realize that I have to go and talk about what I liked so much about the first-generation Pebble Steel (which I'll just call the 'original' from now on). The original had a lot going for it, but the main selling point that I thought about when I first talked about the thing was the way that it enhanced my quality of life. The thing that it brought to me that my Timexes (Timices?) that came before it did not was that it gave me more control over how long I interacted with the computers in my life, and allowed me to keep my head up and out of my phone more of the time. In some senses, what I liked about it is that it felt unobtrusive: it didn't have a particularly "shouty" appearance on my wrist, and the software seemed to go out of its way to get out of mine. This is, I think, what defined what made the original so great for me, and why I liked it so much.

Let's get back to talking about the Time, though. I'll begin with the hardware, because it's ultimately pretty easy to review: there are few opinions, and mostly facts. The one-sentence summary is that the hardware is different in ways that are probably the right thing (even if I don't love them), and there were a few bone-headed mistakes that probably could have been fixed if they had another six months to get to market, but aren't bad enough to completely torpedo the device. The industrial design has changed with the new devices, and this time, the non-Steel Time doesn't look like a plastic toy; the Steel version, which I have, looks only slightly different from the non-Steel one. The Time Steel feels quite a lot larger on my wrist than the original Steel; it doesn't stick out more, but it's enough wider, and the aspect ratio is squarer, that it has the "black monolith" appearance. This is not something that I love, but on the other hand, their hand may have been forced here: the competition, Apple, just released a black-monolith-style device, and there's probably a decent argument that if you want to play in the smartwatch space right now, you need to look like what Apple has set the market up to look like. (The device has the advantage over Apple's that the screen is on all the time, which takes it from "useless monolith" to "actually functional", though.)

The Time tries to make itself known in other ways, too. The steel clasp that came with Kickstarter orders, for instance, is substantially wider than the steel clasp from the original, and has a less anonymous look. (I find the steel clasp sufficiently offensive that I use only the leather strap with mine, now.) Again, though, this is a design point that is not necessarily good or bad on the whole; business-wise, it's probably the right decision to make the high-end version of their watch stand out, but it's not what I'm personally looking for to wear around. On the plus side, the bands are now standard 22mm watch bands, which means that if I find something I like better, I can use it; the original Steel has a proprietary band that was non-trivial to find replacements for.

There were a few silly mistakes in the construction of the device. For one, they used Gorilla Glass for the lens, but used a non-scratch-resistant anti-reflective coating, which meant that all the wonderful scratch resistance of the lens itself is for naught. Empirically, the lens on the Time has held up less well than that of my original; wearing it on my climbing harness for a day or two (as I did with my original at some point), I managed to put a pretty substantial gouge even in the glass of the lens, and a few nasty scrapes on the coating. One corner of the lens even seems to be cracked, which is surprising, because at no point did the device take any terribly substantial hits. By way of comparison, there are some scratches in the lens on my original, but they do not make themselves as apparent as those on my Time. And in a similar vein, the coating on the stainless steel surrounding the lens is, as I understand it, vapor-deposition; this means that it's not terribly sturdy, and even light brushes against certain surfaces can scratch the black coating off of the steel, leaving you with steel showing through. Hard-anodized aluminum probably seems like it would have been a better choice.

Paradoxically, I actually don't mind this property, though. I generally believe that "objects are meant to be used", and an object that is pristine and shiny is an object that hasn't actually been lived with and used over time. Having a few scratches in my watch is something that shows that I actually use it. Some people might find this unpleasant, though.

On a final note on the hardware, I should comment on the battery life. The battery life on the Time Steel, simply put, is fantastic. I got it over a week ago, and charged it once on Friday evening; it's now the next Sunday evening, and I still have 5% of a tank left (which has lasted me all day). I imagine that I'll get into tomorrow before it goes into 'emergency power save' mode, in which it just reverts to showing the time, and then easily be able to get back home before it completely runs out. This is better than the original by a few days, at the very least; additionally, this is with a "connected" watchface that periodically goes and updates the weather.

So that's the hardware for the device. Some neutral changes (I dislike them, but I can understand how others would like them); some stupid mistakes; and some real improvements.

Now let's talk about the software.

Writing a review of the software is hard, because the people -- the real human beings -- at Pebble obviously poured a lot of effort and love into it. This reflects in a lot of ways, but I think the most powerful is the expression of joy that pervades the interface; there's a sense of lightheartedness in the metaphor of a cup of coffee brewing for a battery charging, and in the animation of a phone jumping all around the screen as the device wants to tell you that someone is calling.

And that's what makes it hard, because being honest here is going to be painful to the people who worked so hard on the software. As a whole, it feels like the software on the Time is a regression from the original device, and I think that the root of the issue is that while everyone was having so much fun with it, nobody stopped to think of why people loved the original device, and how they could provide functional improvements to it. I'll provide a couple examples of UI regressions, comparatively, where I can.

I think the regression that is a microcosm for the rest of them is the addition of animations. The Time adds whimsical animations between almost everything that you do on the system. For instance, switching from the watchface to the menu is a folding-and-unfolding animation that takes in excess of a second -- an eternity when all I want to do is pause my music, or start a timer. On the original, I could set a timer with the device responding as fast as I could push the buttons -- center for the menu, down for the timer app, center to start the timer that I had loaded. Changing back to the watchface, or into the new timeline, is similarly slow, and the device ignores input while it's busy transitioning.

This is a fundamental change from the old Pebble experience, in which the device goes out of its way to get out of yours. The Time seems to want to be looked at, under the theory that the experience is what enriches, rather than what you're trying to do. Obviously, I disagree. An interesting study to perform would be to know how many times users press buttons during animations on the original, and on the Time; I would hazard to guess that the rate is quite a lot higher on the Time.

The obsession with the experience extends to the other valuable feature of the Pebble, as well. A usage pattern of the original that I really enjoyed was that my pocket would buzz, my wrist would buzz, and then I'd glance down as soon as it buzzed to see what happened, and whether I cared. The Time, sadly, makes this pattern much more difficult: every time a notification comes in, a picture of the notifying application (a little Facebook Messenger logo, or a SMS logo) is elaborately painted on screen, and then shrunk out of the way, and then, at long last, the actual message is shown. Even more frustratingly, a third of the screen is still taken up by an icon, and not by text; with a long sender name, this might mean that there's no message text at all on the screen, and I'm forced to scroll.

This doesn't mean that there isn't room for whimsy and style, of course. But, functionality has to come first. Having colorful logos with vector animations that sparkle and shine are completely reasonable, and there can be room for that somewhere on the screen, but not at the expense of most of the screen real estate. Whimsy could be used to enhance the experience, instead of detracting from it, too: for instance, special senders could be highlighted in different ways, and given different icons, allowing the user to take in more information while spending less time reading. But once again, this comes down to a focus on experience for the sake of experience, not experience for the sake of enrichment.

In this family, there are a few other design defects. The timeline takes too long to scroll through, because it's animating, and doesn't put as much information on screen as it could. (Maybe color would be useful for weather?) The main menu is completely incoherent, keeping only one item on screen at a time, and filling the screen with empty space; sometimes button presses are dropped or take a long time, so navigating the application menu without looking at it is impossible. (On the original, I know that back-back-back-center-down-down-center will always get me to a BPM counter app, and I could do it on a whim just to check something out. I mostly just don't use the BPM counter app on my Time, because it's such a pain to get to.) Changing watchfaces takes way too many button presses, and the main menu is not optimized for that in general; some options should probably be grouped together (stopwatch and timer, for instance, probably should just be linked together, like the Multi Timer app does). And there are some things that are just inconsistent; the new settings app is a menu unlike anything else on the system (which, paradoxically, is probably the most usable menu I've come across on the device yet).

The design defects are the things that worry me the most, because somebody may have believed that they were right to begin with; I suspect that a user study would have borne out that if the goal is to minimize time interacting with the watch (and frustration with the experience), then these design decisions may not be optimal. There are, of course, as with any new product, a handful of other bugs. They don't worry me as much, because they are unambiguously bugs, but for completeness, I may as well list them. Many of them feel like issues that come as a result of having simply run out of development time at some point.

The major one is probably that the Pebble app has been split into two -- one for the original Pebble, and one for the Time. There are probably good technical reasons for this, but the effect is still that I have two apps installed on my phone, and that the app that I already had didn't work with my watch on the first day I unwrapped it. (The old Pebble app, in fact, tried to talk to my new watch, and provided no warning that it couldn't help me; it got as far as trying to update the firmware on the new watch, and then aborted, saying that no firmware could be found for it.)

I suspect that this is the cause of another regression; the XMPP client I use, Conversations, supports Pebble natively (rather than with notification-scraping), but since I started using the Time, the Pebble support in Conversations no longer seems to function, and I get notifications sent as if they were scraped, instead. I don't know if this is because I have both Pebble apps installed, or if it's because Conversations is using an old Pebble API, but either way, it used to work, and it doesn't now.

Finally, the voice support is generally good, but occasionally buggy. For instance, after I activate it, I sometimes have to wait quite a while for something (I don't know what) to happen before I can begin speaking; this makes it far less useful, since it means that I have to spend much more time looking down at my watch (especially bad if I'm driving, where it almost entirely defeats the point). Similarly, sometimes voice recognition simply fails, and I have to try again. I don't know why the watch can't try again for me. I think the most pressing issue, though, is that having a conversation with voice is impossible because of a different bug: if a notification comes in while you are dictating, that notification will be lost forever (well, or until you navigate the tedious menus to find the notification list) when you leave the dictation screen. That means that if I'm having a conversation with someone, and I'm responding, and they say something while I'm responding (or trying again, since the voice system inexplicably failed), then the conversation is effectively over, because I've dropped a message and even if I dig it up, I can't respond. The plus side is that the recognition quality is quite good when it works; I wish it could infer punctuation from tone, pauses, and grammar, but I imagine I'll get used to that.

So that's the software situation. I have confidence that they'll eventually fix the little bugs. The design bugs, however, still scare the pants off of me.

My feelings about this thing are somewhat more mixed than the original Pebble. A year and a half ago, I wrote:

[Smartwatches] seem like they have a real, genuine possibility to enrich. This one differentiates itself from the others right now because it asks not to be used -- and maybe that's why it's so charming.

Many of the things that I loved about the original -- the fact that it was small, tried to be anonymous, and tried to get out of your way -- are lost in the Time. The Time seems like an evolution of Pebble, to be sure... but I'm not sure that it's the evolution for me. The consolation, I suppose, is that there are a lot of smart people at Pebble, who very obviously deeply care about their product. I think that some errors were made in the design so far, but I still have confidence in their ability to get it right; to differentiate themselves by building a product that genuinely enriches the lives of lots of people; and return to their roots of why Pebble was such a great product to begin with.

(buy PVC)

My Website Powered by LiveJournal.com