Friday, March 27, 2015

Part 3 of Who are you designing for? Or, thoughts on the Lenovo ThinkPad X1 Carbon. Or, trying to buy some more happiness.

Let's talk about the incredible decision to radically change the keyboard of the ThinkPad X1 Carbon 2nd Generation, when a great, traditional, and high-performing keyboard has been the defining characteristic of the ThinkPad brand.



The first big change is that the mouse buttons are gone.  The second big change is that the physical function keys are replaced with a touch strip.  The third big change is that some of the keys are re-arranged from a more traditional arrangement.  The first two changes are disastrous; the third is just a bad idea that I actually kind of like.

Instead of physical mouse buttons, the entire touchpad clicks, and the system then guesses which button you intended to press based on where you are touching it with your fingers, where your fingers are near its surface, where the thumbs you didn't realize count as fingers are, and the purity of your heart.  This does not work.  I got the mouse click I wanted about 50% of the time, and sometimes needed three or four or five clicks to get the desired result.  If this were a mouse I would consider it broken.  I suppose results may be different if you use the touchpad instead of the TrackPoint as the pointer, and perhaps none of the human beings who ever physically touched an X1C2 prototype ever used the TrackPoint; it's hard to imagine any other scenario in which this was not considered a showstopper problem.  The ThinkPad design blog offers this justification:
We unified the clickpad by integrating the trackpoint buttons into the elegant glass touchpad, making it appear even larger and more streamline.
The second bad decision is the function key touchstrip.  It has no tactile response to clicking, it has no tactile way to know where the key boundaries are, it takes around a second to change modes, and it is sensitive to false positives.  Each and every one of these four problems individually disqualifies it from replacing physical keys.  It offers no significant benefit over keys in usability.  I can't even speculate why they thought it was a good idea.  Here is what it's like to use to lower the volume:

1. look at it
2. touch it in the general region of the volume down icon
3. wait a moment while nothing happens
4. realize it's in the wrong mode.  If it was in the right mode before, you probably missed it and hit the adjacent mode-changing button anyway, but it was probably already in the wrong mode.
5. touch the mode-changing button (Linux only supports two of the four modes, so I guess that's a blessing in disguise)
5. wait a moment while nothing happens
6. look at it to confirm it's changed modes
7. touch it in the general region of the volume down button
8. wait a moment while nothing happens

and then the volume changes.

Here are the steps with a physical key, after your body has integrated it into your muscle memory:

1. press the volume down key

Even people who don't touch-type probably develop this muscle memory quite easily, I speculate without any evidence.  The touch strip is a horrible design decision.  Instead of working function keys and laptop control keys, the X1C2 has a replacement that is worse in every way and better in none.  For anyone who routinely uses function keys in touch-typing, this should render the X1C2 unusable; for me it was a close call.  I can't really imagine which user they thought would benefit.  It's not like they don't test these things:
... to determine if and how we would make the changes to our keyboard in 2012, we embarked on one of the most in-depth keyboard studies ever conducted for ThinkPad. We did 350 hours of user testing with people in four countries. With each participant, we conducted 90- to 120-minute one-on-one interviews with hands-on use of different keyboard conditions to understand the latest about keyboard use and design preferences.
What doesn't make sense is how they managed to put that much effort into the new keyboard redesign in the X1 Carbon 1st Generation, which had a normal function key row and normal mouse buttons, and then so completely screw up the 2nd Generation.  There are mysterious hints of trouble at Yamato Labs in these interviews with the designers. These comments are from 2012, predating the X1 2nd Generation, but they do not instill confidence in the stewards of the brand.
I heard that there was a major turnaround during development. Did it involve the mechanical engineering?
Otsuka: Yes. I believe it affected the mechanical engineering more than anything else.

There were significant changes after the injection mold and die cast mold were completed. Of course, the changes also affected the product direction and the project as a whole, so all hell broke loose at the Yamato Office. Opinions on the changes were divided even within the company.

However, we took on the challenge that this major turnaround presented.

...

Looking back now I believe it was the right decision, as the X1 Carbon was improved so much that it became almost a completely different product to the initial plan.

I feel like for this one product we did enough design work for two or three products. I'm exhausted (LOL).
And:

Otsuka: Allow me to share three surprises I think the whole development team would agree we encountered for the X1 Carbon.

1. The fact we actually achieved our goals
2. The changes in key members
3. A major turnaround

Unfortunately I can't go into the details of the changes in key members or the major turnaround.
And:
A message from Akira Fukushima
This is Akira Fukushima, the new head of ThinkPad development Yamato Labs since August of this year [2013].
Would you trust this man with your eyes and hands?

It's fascinating that the blog provides all of this detail, and certainly the fact that I'm poring through it illustrates an alarming, almost Apple-fanatic-esque depth of investment on my part in the details of this piece of consumer electronics.  All I can say is that the moral has to be, don't be weird and picky, or nothing will ever be to your satisfaction, certainly nothing that requires economies of scale.

But let me pretend to sociological/marketing/usability research and point out that, despite all of this information, there is no good explanation of what benefits they anticipated from these two horrible changes. More hints in these comments:
Why, oh why do you remove the seventh row of the keys? Because it "avoids visual complexity and most people have no idea what these keys even do"? That's what you say in one of your videos. (Dom Delimar)
I took the liberty of calling 2 people at Morrisville and Raleigh and finding out what the reason was behind this and the answer was quite interesting: Lenovo are selling more of this type of keyboard. So that means die-hard ThinkPad enthusiasts are now becoming a minority in the sales figures and new Lenovo customers who probably have never had the history of typing on ThinkPads for decades, are on the rise. (ElliotR)
That's a pretty tough spot for brand stewards. How do you keep the die-hard fans and also grow your market? We'll come back to that after we talk about the third major, controversial change to the keyboard. Have you spotted it yet?

No comments :

Post a Comment