That’s how I noticed this new review of FatWatch, my weight tracking app for iPhone:

It does exactly what it says it does, but it’s far too expensive and the developer seems to have abandoned it. Get one of the free weight apps and save yourself the dough.

Ouch!

In a world of disposable mobile apps, FatWatch might seem expensive, but it’s actually quite cheap for what it does: tracking your weight against a moving average in a well-designed application. The average is key; it’s the only useful way to track a human being’s weight, and none of the free apps can do the math for you.

For devotees of The Hacker’s Diet, it also lets you import your weight history from the old Palm Eat Watch app and export it to your computer any time you choose.

As for the other concern, I assure you that FatWatch has not been abandoned. I use it every day! I suppose it is overdue for an update, but that’s only to add new features, as (save for a minor cosmetic issue) no bugs have been reported in FatWatch 1.4.

The good news is that I recently completed a contract project that puts me in a comfortable enough financial position to devote time to a FatWatch update. So, watch this space, and if you’re interested in beta testing, wait for an announcement soon.

I submitted an update to Apple, changing as little else as possible, hoping that it would get through their review process in time for the public release of 3.0. Unfortunately, it was rejected for violating the iPhone Human Interface Guidelines. That’s right: I was nailed for something that they previously approved.

What I got away with

I could bitch about it, but they were right: the objection was to the Wi-Fi Import/Export feature, which in FatWatch 1.4 was enabled by a simple switch.

If you want to import or export a CSV file containing your weight history, all you had to do was turn that switch, and a new row would appear beneath it with an IP address to connect to.

If you didn’t know what to do with all those geeky numbers and slashes, hopefully you would tap on it, because doing so would display an alert explaining that you are supposed to enter that into the address bar of a web browser on your computer.

Even when working properly, this is a spartan interface. I’d honestly be surprised if anybody who doesn’t describe themselves as a “geek” figured it out. What’s worse (and this is where I got into trouble with Apple) is when you turn on Wi-Fi Import/Export and you’re not even connected to a Wi-Fi network. You get a blank row.

So yes, I could say to Apple, but that’s how it worked all along! But that’s like telling the police officer you shouldn’t get a ticket because you run this red light every day and nobody gave you a ticket before.

Making things right

Initially, I tried to fix things by detecting whether a Wi-Fi connection was available and, if not, snapping the switch back to the OFF position and displaying an alert with an error message. While that would have been an improvement, it still felt… terrible.

After sleeping on it (always more productive than hammering away all night, will I ever learn?) I realized that Wi-Fi Import/Export really deserves its own screen, rather than the series of alert views I had allocated it. In practice, you aren’t going to use your iPhone or iPod touch for anything else while you are importing or exporting a file, so there’s no need for it to operate in the “background” of the app.

So in FatWatch 1.4.1, the Wi-Fi Import/Export changes from a switch to a menu item.

When you tap it, you see a screen that explains what Wi-Fi Import/Export does and how to use it. It also tells you that you can use Bonjour, instead of typing in a series of dots and colons and numbers. FatWatch always used Bonjour, but there wasn’t enough space in the old interface to explain it.

FatWatch will now remember the last time you imported or exported a file; very useful if you want to export a file as a backup every so often and you can’t remember when you did it last.

But, most importantly, if you are not connected to a Wi-Fi network, FatWatch now makes it abundantly clear.

As a bonus, now that I have a whole screen for this feature, I was able to get rid of the alerts and do some nice things, like display a progress bar so that you know things are working when you’re importing a large file.

I won’t claim it’s perfect: even now I want to edit some of the text in those screenshots. I held myself to a fairly tight schedule because I didn’t want people to be waiting too long for the bug fixes that was supposed to be the primary focus of this release. However, I am glad Apple challenged me to improve this feature, because it has gone from an embarrassment to something I can feel mildly proud of.

More

A few other improvements that are coming in 1.4.1:

• The Wi-Fi Import/Export web pages are easier on the eyes.
• FatWatch puts today’s date in the exported file name, for example: weight-2009-06-29.csv. Useful if you let those files accumulate in your downloads folder.
• The Go To Date screen displays its buttons in a navigation bar; looks a lot nicer.
• Fixed a bug where BMI values were calculated incorrectly. Basically, if you select your height in feet and inches, FatWatch always added an extra centimeter to your height. That has been fixed.

Everybody cross your fingers, and hopefully it will appear on iTunes soon!

If you are using FatWatch 1.4 (the latest version at the time of this writing) then you can safely upgrade to iPhone OS 3.0 if you are willing to suffer a cosmetic issue. Specifically, if you have BMI Monitoring enabled and use the control on the top of the Log view to switch between BMI and Variance, the values will not update unless you force them to by either scrolling the table or switching to and back from another view. It’s annoying, but not fatal.

I submitted an update to Apple that fixes this issue (and a few others), hoping it would be available on iTunes before the 3.0 update was released. Unfortunately, Apple rejected the update for a human interface issue concerning a feature that I hadn’t changed. It feels a little unfair to be rejected for something that was previously approved, but “I got away with it before” is not really a defense. They are right, it is something I should fix, it’s just that the timing is obviously less than ideal.

I intend to fix this issue and resubmit today or tomorrow, which means that FatWatch 1.4.1 should hopefully be available sometime next week. Or maybe not. I wish I could say. In the meantime, should you want to install the iPhone OS 3.0 update you can continue to use FatWatch 1.4 and be mildly annoyed at the display issue.

While you’re waiting for that to become available and/or affordable, you might be interested in FatWatch, an app for your iPhone that also estimates your caloric intake, although it requires you to step on a scale and make a note of the number once a day.

(P.S. I found the above news via The Skeptical Hypochondriac. Recommended if you’re interested in health news without the hysteria.)

First, you may now rate an application without writing a review. Previously, you had to write something, unless you were using Apple’s “rate on delete” feature. I hope this will restore some balance by making it easier for people in a hurry to provide a positive rating, although it is still up to developers to direct satisfied customers to the app’s review page. ;-)

Second, and I think this is a big deal, reviews are now tagged with the version of the software being reviewed. Average ratings are computed for both the current version and across all versions. Thanks to this change, I was able to compile following table for FatWatch:

Version	★★★★★	★★★★	★★★	★★	★	Average
1.4	3	1				4.75
1.3	6	1	1			4.63
1.1	2	3			2	3.43
1.0			1		2	1.67
Total	11	5	2		4	3.86

(In case you think I’m cooking the books, version 1.2 was never released to the public.)

iTunes reports that FatWatch’s average rating of 3 stars is based on 50 ratings but only 22 reviews. With some algebra I can determine that the remaining 28 reviews have an average of 2.32 stars. If those are all from “rate on delete,” that’s seems pretty good.

I think this paints a more favorable (and honest) picture: FatWatch got off to a rough start, but its grades have been steadily improving. I’m relieved, because two of the negative reviews about version 1.0 were voted as Most Helpful and used to appear on the main page.

I’m happy that Apple has improved the App Store in this way. A lot of developers are quick to criticize them, but I know they aren’t out to get me, and that running a huge operation like the iTunes Store isn’t easy.

Anyway, you’re probably wondering what I’m up to and when the next update to FatWatch will be released. I have been dividing my time between FatWatch and consulting, and lately the divide has not been as even as I’d like. I have several pages of ideas sketched out, however, and I’m eager to work on them. The next update will probably focus on improvements to the Trends feature, specifically on making the calories per day number easier to interpret and apply to your goals.

Anyway, if you’re curious about what’s changed since FatWatch 1.3, here is a complete list of changes in FatWatch 1.4, minus some under the hood stuff that is only interesting to me.

Big New Feature: Body Mass Index (BMI) Monitoring

To enable, go to the More tab, switch it on, and select your height. Afterward BMI will be displayed in various places around the app, using the following color scheme: red for obese, yellow for overweight, green for normal, and blue for underweight. Classifications are based on the World Health Organization’s guidelines. Think BMI is bunk? Don’t turn it on, and you won’t have to look at it.

Log

• Added a control to toggle between displaying Variance (your scale weight’s distance from the moving average) or your BMI (if enabled).
• Improved drawing of long notes.
• Added a “hint” cell at the top of the table.
• “Go To” screen has larger buttons.

Weigh-in

• Visual update, changed background color and tweaked control layout.
• If BMI is enabled, weight picker cells are colored according to BMI zone.
• Note field is larger (easier to tap), supports multiple lines of text, and has sexy rounded corners.

Trends

• Added overeating and undereating labels to the energy surplus/deficit field.
• Added total weight gained or lost during trend period.

Goal

• Goal weight picker uses whole numbers, regardless of selected scale precision.
• If BMI is enabled, start BMI is displayed, and you can set a goal by selecting a BMI value in addition to selecting a weight.
• Added an info button to explain how start weight is computed (its the moving average, not your scale weight).

More

• Added a switch to enable BMI monitoring.
• Added a button to open FatWatch’s page in the App Store.
• Removed unnecessary section titles and added a hint referring to the Settings app.

Chart

• Now offers several preset zoom levels: Past Month, Past Quarter, Past Year, All Time, and Browse (the original scrollable view).
• If BMI is enabled, chart regions are shaded to indicate BMI zone.
• If database is empty, displays an empty chart (instead of a boring-to-look-at message).
• Browse mode scrolling is linked to Log: start position is determined by current Log position.
• Label text is drawn using Unicode encoding, making better international support.

Settings

• Renamed “Scale Increment” to “Scale Precision” — the latter sounds more precise, don’t you think?
• Added a switch to disable Bonjour, in case you don’t want to publish Wi-Fi Import/Export’s web address when you turn it on.
• Shade Weekends is now enabled by default.
• Added “Shade BMI Zones” in case you don’t like so much color in your chart.

Miscellaneous

• Added a startup image.
• When you quit FatWatch, it remembers which tab you had selected (i.e., Log, Trends, Goal, or More) and returns to it when you next launch the app.
• Setup screen respects existing preferences.
• Added a cancel button to the Set Passcode screen (that never should have been left out in the first place).
• Wi-Fi Import/Export displays device name (“iPhone” vs. “iPod touch”) as appropriate.

I hope you find it worth the wait.

When you launch FatWatch, most of the time you probably want to quickly enter your weight and then close it immediately, before getting dressed and heading to work. I decided to design the weigh-in screen to be as efficient as possible in this scenario. When considering whether to “split the picker”, there are two considerations that work in the single picker’s favor: scrolling distance and Fitt’s Law.

It is true that scrolling from FatWatch’s default of 200 lbs to your initial weight can be tedious, especially if you’ve selected a scale precision of 0.1 or less. However, from that point on, FatWatch remembers your weight and always defaults to your average weight on the day you select. Since your scale reading will only be a few units away from your average, you won’t have far to scroll.

An example of a split picker.

If each digit is on a separate wheel, the amount of scrolling required varies wildly depending on the numbers involved. For example, changing from 168.5 to 167.5 is easy: adjust the third wheel one unit. But to change from 168.5 to 170.0: adjust the second wheel one unit up, the third wheel eight units down, and the last wheel 5 units down. Got that? It’s not hard to figure out, but it is work that your brain has to do. And it has to do it early in the morning when you’d rather be crawling back into bed.

FatWatch's weight picker.

Now, all that work is worth doing if we don’t know anything about the number you are going to pick. When you set an alarm, for example, you are just as likely to pick 8:25 AM as 11:00 PM, so an alarm clock app is probably saving you work by letting you pick the hour and minute separately. But since FatWatch knows the number you want to enter won’t be far from the default, a single picker makes more sense.

And what about Fitts’s Law? It’s a principle of interface design that states the time to acquire a target is inversely proportional to its size. In other words, it takes more time to tap a small thing than a big thing. By using one wheel, FatWatch gives you a very large target to hit and swipe with your finger. If it was split into three or four wheels, you would have four separate and much narrower targets. So, not only do you have more wheels to adjust, it will take you longer to adjust each one of them.

I hope you enjoyed this explanation of the thinking behind FatWatch’s design. The 1.4 update is almost finished; when it’s done I’ll have more design stuff to write about.

I’m also going to use curse words.

Somebody using the pseudonym Overweight on the Joel on Software message board started a thread about his New Year’s resolution to lose some weight and wondered if they could apply Evidence Based Scheduling, a project management method, to the problem. The very first response, in its entirety, was:

This may sound harsh, but…

all the software in the world won’t help unless you do the two simple things: eat less, exercise more.

Once you find you can do that consistently under your own power for a couple of weeks, you can look for software support. But honestly, losing weight is just that easy.

farmboy
Monday, January 05, 2009

farmboy, fuck you.

The attitude behind this “helpful” advice is the sort of thing that causes so many overweight people to feel guilty and miserable most of the time. farmboy, you literally say that losing weight is easy while simultaneously accusing Overweight of avoiding the (easy?) work by trying to solve it with software.

Guess what? Most people who are concerned about their weight are (1) not morons and (2) very much aware of the benefits of eating less and exercising more. They are overweight because they have trouble doing those things, and while lots of people will smugly tell them that’s all they need to do, there are fewer people offering useful advice on how to do them.

Let me make an analogy. Let’s say somebody posted a thread about being in debt, asking for recommendations of finance software. What would you say to a person who responded:

all the software in the world won’t help unless you do the two simple things: spend less, save more.

Once you find you can do that consistently under your own power for a couple of weeks, you can look for software support. But honestly, getting out of debt is just that easy.

That’s fucking crazy! Everybody will tell you the first thing you do when trying to get your finances under control is to get out a sheet of paper, track your income and expenses, and make a budget. It’s simply assumed that you need to be watching the numbers before you can control them. Maybe people would think of weight management the same way if your ass mailed you a monthly statement just like your bank did.

Some people don’t use any software and rarely overdraft their checking account because they have a pretty good sense of how much money is in there. That’s just how they are wired. Good for them. Some people do not have that sense, so we tell them to use Quicken, or Mint, or iBank, or MoneyWell, or a sheet of paper, to give them that sense.

It’s the same way with weight. Some people don’t think about their weight and unconsciously adjust their eating habits accordingly. Some people, like myself, don’t have that sense: our appetite is totally unconnected from the rest of our bodies. I feel a need to eat even when I know I shouldn’t. Your smug lecture about how easy it is to lose weight doesn’t really help at those times.

So I use a weight tracker. I step on the scale every morning and record it. The software does a little bit of math so I can understand the numbers better. It is a prosthetic, giving me a sense of my weight that other people have naturally.

Nobody is saying that tracking alone will cause you to lose weight (or save money), but oh my goodness it makes such a huge difference that you probably will. Simply having that awareness, possibly for the first time in your life, will influence all of the tiny decisions about food that you make throughout the day, and it actually will have a small effect.

And when you start going to the gym on a regular basis, and nobody has commented that you look any different, and your pants basically fit the same, and mostly you’re just more sore than you used to be, oh my goodness it is so awesome to see the trend line on the chart is pointing downward. And maybe you’ll keep at it for a week or two more.

So, if you’re overweight and want a resolution for 2009, here’s what I recommend: start tracking your weight. Don’t worry about how you’re going to lose it yet, just start tracking it. Ideally, step on a scale every day and enter the number into some software like FatWatch. Once you have a sense of your weight, you’ll have the power to get it under control.

It’s really a clever game (I can say this because, while I did the programming, the idea was not mine), with just the right combination of luck and skill required. You roll colored dice and then arrange them to get points depending on the probability of the combinations you create. It’s easier to understand if you watch the instructional video on YouTube.

Lock ‘n’ Roll will be free for a limited time, so go download it now before we decide to start making money.