I’m : a programmer, writer, podcaster, geek, and coffee enthusiast.

Photo check deposits

Almost two years ago, Chase updated its iPhone app to include photo check deposits:

Deposit checks with two camera clicks.

Instead of driving to the bank, you can deposit your check with your Chase Mobile App. Just snap a picture of the front and back of your endorsed check and send it using your Chase Mobile App.

This is one of those ideas that sounded great until I actually tried it. In practice, here’s how this works:

  1. Log into the app, which means typing at least your long, complex banking password, and probably also your long, complex username, on the iPhone keyboard.
  2. Navigate to the deposit screen. Pick the target account. (The interface is very navigation-heavy, requiring numerous unnecessary taps throughout the whole process.)
  3. Type in the check amount manually.
  4. Photograph the front of the check. Chase recommends taking the picture in good lighting on a dark, non-reflective surface while standing up. They also caution you to ensure that it’s not blurry, and if any of these conditions aren’t met, it might not be accepted.
  5. Photograph the back (endorsed, of course) with the same precision and warnings.
  6. Submit it.

I’ve tried photo-depositing six checks. Two of them failed that last step with “Unable to connect to Chase, try again later” messages. When that happens, all of the fields and photos are cleared and you need to start over.

Did you want to deposit multiple checks? That’s multiple separate deposits. There are photo-deposit limits of $2,000 per day and $5,000 per month, neither of which the app tells you until you attempt to cross those limits.

If I have a few checks to deposit, and any one of them is refused for photo deposit, I need to go to the bank anyway, removing the convenience that the app was supposed to provide.

But assuming that a check successfully gets submitted, then what? Simply being submitted doesn’t mean that the check has been accepted. You need to wait a few days to see if the check posts to your account before you know whether it has been accepted. And then you’re supposed to destroy the paper check.

So I need to keep this check somewhere, remember not to deposit it again, and remember to check my account in a few days to ensure that it clears. This is a physical and mental organizational burden: this check effectively stays in my mental “incomplete” box, and sitting on my desk, until it clears, which I probably won’t be notified about.

And then, after it clears, I’m told to rip it up and throw it away, which goes against all of my instincts as a responsible person.

My alternative is depositing checks into an ATM, which is much simpler:

  1. Go to the ATM. (The pain-in-the-ass factor on this varies per person, but I end up walking past a Chase ATM regularly.)
  2. Enter a 4-digit PIN, tap Deposit, tap checking account.
  3. Insert a stack of checks.
  4. Wait about 10 seconds for it to automatically scan them and recognize their amounts. I rarely need to manually enter a check’s amount, even when it’s handwritten.
  5. Take my receipt, walk across the street to the deli, and get a sandwich.

It’s much faster and simpler than a photo deposit. (I can also get cash while I’m there. Can’t do that with the iPhone app.)

And then it’s done. The check is out of sight and out of mind. I know that if anything goes wrong, the bank will mail me something about it, although I’ve never had an ATM-deposited check get rejected by the bank later.

Sometimes, new technology is not progress.

Four LED light bulbs reviewed for normal use

Generally, I hate CFLs.

They’re big, they’re complicated, they’re ugly, they’re harsh to look at, they usually need a warm-up period before reaching full brightness and proper color, they contain traces of (highly toxic) mercury, they’re fragile, and they don’t work well on dimmers. And they often stop being useful long before their stated lifespans because their color shifts or they lose brightness.

But they use much less energy than incandescents, they’re (now) inexpensive, and they’re a great way to get much more light out of a fixture than its maximum wattage previously allowed — as long as you don’t mind waiting for the warm-up. So for certain uses, they can be great. But they’re a poor choice for a lot of household roles, including any exposed bulbs (too harsh, too ugly) and any area that’s usually only lit momentarily such as bathrooms, hallways, and closets (not enough time to warm up).

LEDs share some of the downsides of CFLs. They’re often bigger, uglier, and more harsh-looking than what they’re replacing, and they don’t dim very well. They aren’t bright enough to replace high-wattage bulbs yet, and they’re very expensive up front. But compared to CFLs, they use much less energy, they’re simpler, they don’t need time to warm up, they’re much more durable, and they don’t contain mercury.

I wanted to try LED bulbs, but my Amazon research only found divided opinions and a lot of uncertainty. With no clear “winner”, I decided to buy four of the top-rated choices fitting my criteria, figuring I’d buy more of whichever I liked best and find tucked-away homes (closets, the utility room, the basement) for the crappy ones.

I also wanted to somewhat objectively compare the LED bulbs to non-LEDs. Here’s the test set:

My setup is fairly reproducible if you want to compare my photos with other bulbs: I photographed them in an otherwise dark room against a white wall at 100mm, ISO 400, f/4, 1/100th with white balance set to 5000K. I put a piece of cardboard behind them for the “Off” pictures to make them more visible against the wall.

(This is a Javascript image switcher. View on the full site to see and compare the bulb photos.)

Off: 60WHalogenGE LEDPhilips LEDG7 LEDEarthLEDCFL
On: 60WHalogenGE LEDPhilips LEDG7 LEDEarthLEDCFL
Lamp: 60WHalogenGE LEDPhilips LEDG7 LEDEarthLEDCFL

Philips EcoVantage halogen-incandescents

These are great. We’ve used them around the house for over a year in various fixtures, and they really shine (no pun intended) in applications where CFLs or LEDs would be too ugly or harsh-looking, such as an exposed downlight fixture like those often found above bathroom sinks. (That’s where we use three of them.)

They’re available everywhere, and they’re visually and functionally almost indistinguishable from normal incandescents. They’ll work almost anywhere an incandescent would work. The “Natural Light” was too cool-colored for most of our uses, but the “Soft White” is perfect almost everywhere in the house. And if the U.S. bans traditional incandescent bulbs, these save enough energy to still be permitted under current regulations.

But they don’t last very long by today’s standards: expect to get about a year out of them. And while they’re more expensive than incandescents, they don’t save nearly as much energy as CFLs or LEDs.

GE Energy Smart LED

I love this bulb. It outputs omnidirectional light, it’s almost as small as regular incandescents, and it’s one of the most normal-looking bulbs in the batch. It’s even nice enough that I’m using it for a couple of partially-exposed bathroom downlights.

But it’s only bright enough to replace approximately a 45W incandescent, and at about $35 each, it’s not a good value for that amount of light. Its color, while not as cool as “daylight” bulbs, is noticeably cooler than “warm white”. I estimate that it’s around 3000K. I love the color, but if you prefer warm, yellowish light, you might not like it.

Philips AmbientLED

A lot of people really love this futuristic-looking yellow bulb. I expected greatness from it, but I was disappointed.

It’s one of the best bulbs I’ve tested for brightness, omnidirectional light, size, shape, and value. But the color is off: in an attempt to match the subtle yellowish tint of standard “soft white” incandescents, it has gone a bit too far, which the photos clearly show when compared to the incandescent.

The AmbientLED is distractingly yellow, almost orange, and it casts subtle, strange, almost pink tint on nearby objects. Rather than looking warm and natural, it looks artificial, like the sucralose of warm-white lighting. I generally like cooler-colored light, but even my wife who prefers warm light disliked the AmbientLED’s color.

G7 Power LED

This was mediocre. It produced decent output and color, but it was too large to fit in some of my fixtures, and its light output is extremely directional, which makes it inappropriate for a lot of uses that I tried, such as table lamps, frosted-glass downlights, and ceiling fixtures that align the bulbs horizontally. The effect is particularly visible in the “Lamp” photos: the bottom half of the lampshade remains dark.

It also had the cheapest-feeling body, and it’s the only LED bulb I tested that didn’t have a heatsink-like base design. I’m skeptical that the smooth metal base can dissipate enough heat.

EarthLED EvoLux 2

The EarthLED is like a better-implemented version of the G7 Power LED: it’s smaller and brighter with a better build quality, the color is a bit warmer, and it’s subjectively the best-looking bulb I tested. But it shares all of the G7’s downsides of directional light output.

If I had a good use for a directional LED, I’d choose EarthLED over G7 Power.

Not covered: dimming

Dimmable LEDs (and CFLs) are a huge mess of bad hacks, broken promises, and disappointment.

I haven’t addressed it here because finding dimmable LED or CFL bulbs and compatible dimmer switches is a huge pain in the ass, and it usually doesn’t work very well. If you want to use dimmers, you’re much better off sticking with incandescents and halogens.

If you really want to use energy-saving lights, consider changing your habits (and possibly your lighting arrangement) so that you don’t need dimmers. It’s very unlikely that CFLs or LEDs will work significantly better with dimmers in the future.

Practical picks

I love the GE Energy Smart LED, but it really needs a 60W-equivalent version to be competitive. That said, I’ve already bought a few more of these because I like them so much. They’re great for table and night-stand lamps. If you have two- or three-bulb ceiling fixtures that don’t need a ton of light (and you can handle the high price), you can try these in there too, although it may not be enough light. I put them in a pair of 3-bulb ceiling fixtures in an 18-by-15-foot room, and it’s barely enough light, but these lights are on so often that they’re going to pay for themselves quickly in electricity savings.

I can’t recommend the Philips AmbientLED. The color is just too weird. But Philips is pushing LEDs forward very aggressively, and I bet they’ll have a more broad general-purpose LED lineup in the near future. I’d love to see Philips offer a cooler-colored option.

If you’re considering a directional LED such as the G7 Power or EarthLED options, make sure your intended fixture can handle directional light without looking stupid. After I finished taking all of the photos for this review, I tried finding permanent homes for all of these bulbs around the house. Since they’re so directional, the G7 Power and EarthLED are the only two for which I haven’t found good spots yet.

For exposed bulbs, dimmers, and other applications for which LEDs and CFLs are poorly suited, the Philips EcoVantage halogen-incandescents are great. They’re not quite as energy-wasteful as regular incandescents, and they have almost no other drawbacks or limitations.

Where does this leave CFLs?

Even with LED bulbs maturing, CFLs still have a role: they’re much cheaper than LEDs up front, and they still offer the highest light output options for most fixtures, as long as you can physically fit really big CFL bulbs and don’t need to look directly at them.

But LEDs can now take over in fixtures intended to be used with 40W or 60W bulbs. And once LEDs can practically and economically replace 100W bulbs, CFLs will lose a lot more ground.

I may have purchased my last CFL.

Predicting WWDC dates and the June 11 rumor

WWDC is widely rumored to be the week of June 11 this year, based solely on a “Corporate Meeting” entry on the Moscone Center calendar that has been there since last year’s WWDC.

I bet that’s wrong.

Granted, WWDC is usually labeled “Corporate Meeting” on the Moscone calendar a few months ahead of time. But this was booked almost a year ago, uncharacteristically far in advance. More importantly, have you checked hotel prices for that week? They’re insanely high, and are almost all sold out already. It looks more like something else is going on in town that week, and attendees for whatever is occupying the Moscone Center have already booked the hotels.

UPDATE: Thanks to Brian Stucki for telling me that the U.S. Open is that week, which filled up the hotels.

I’ve also seen a few people suggest that WWDC will be held in the fall, since that’s when the next iPhone will probably be released. That’s possible, although I don’t think it’s likely: Apple successfully severed WWDC from the new-iPhone launch in 2011, and I don’t see why it can’t continue to be a software-only event.

Moreover, looking at the Moscone calendar, availability in the fall is pretty tight. Here’s availability for Moscone West for every week from May through Thanksgiving, as of today:

What if Apple finally convinced Moscone not to put a publicly visible entry on their calendar for the WWDC booking?1

Given the hotel issues and unusual advance booking for the week of June 11, I bet that’s not WWDC.

The more likely weeks are June 4, July 16, July 23, or August 20. But July and August are so far into the summer that they’re more likely to interfere with people’s vacations, both for attendees and Apple’s staff. June is a much better month for this.

So I’m guessing that WWDC 2012 will be the week of June 4. And I hope we don’t need to wait much longer to find out.

  1. A few people have also suggested that maybe Apple would move WWDC to a different venue. Apple’s stated reason for using Moscone West so much recently was because it was the biggest suitable venue in the area. I don’t know the region well enough to know if that has changed, but I’ll assume that WWDC will still be held at Moscone West this year.

    It’s also possible that something else could be different this year. Maybe Apple booked the hotel rooms for us and is including them in the price (very unlikely). Maybe check-in won’t be on a Sunday with the sessions from Monday through Friday. Maybe there won’t even be a WWDC. But as fun as it is to wildly speculate on major changes, it’s usually a pretty safe bet that Apple will continue to conduct their events the way they’ve been conducting them. ↩︎

Euphemisms for death that big companies uncomfortably use to describe someone who quit or got fired, which I always thought was weird

Twitter’s “Innovator’s Patent Agreement”

This is an interesting move by Twitter, possibly becoming public now in response to a lot of recent fears about Twitter’s patent on pull-to-refresh:

[The Innovator’s Patent Agreement] is a commitment from Twitter to our employees that patents can only be used for defensive purposes. We will not use the patents from employees’ inventions in offensive litigation without their permission. What’s more, this control flows with the patents, so if we sold them to others, they could only use them as the inventor intended.

A lot of people are heaping praise upon them for this, but it’s important to maintain perspective.

First, it’s interesting to read their definition of “a Defensive Purpose”, which, loosely summarized, means they can use their patents to:

  1. Countersue anyone who sues or threatens to sue Twitter for any intellectual-property claim.
  2. Sue anyone who has sued anyone else for patent infringement in the last 10 years.
  3. Sue anyone “to deter a patent litigation threat” against Twitter.

These leave a lot of room for interpretation.

The first definition just says “intellectual property”, not just patents. Suppose I implemented pull-to-refresh in Instapaper, Twitter released a new read-things-later product called Instasaver, and I sued them for trademark infringement. They could countersue me for patent infringement on pull-to-refresh.

Or suppose that Twitter implemented tilt scrolling, a lazy designer there just stole my toolbar icon for it, and I merely threatened to sue them for copyright infringement if they didn’t change it. After all, any copyright-infringement claim is at least an implied lawsuit threat. Twitter could then sue me for patent infringement on pull-to-refresh.

The third definition, “to deter a patent litigation threat”, could be interpreted very broadly. Apple has a lot of patents. What if Apple implemented pull-to-refresh in Mail, and a few years down the line, their relationship with Twitter went south for other reasons? Might Twitter conclude that Apple’s patent portfolio could represent a litigation threat, and preemptively sue Apple for pull-to-refresh as a “deterrent”?1

Even if all “Defensive Purposes” were restricted to patent litigation, what if it’s your perfectly valid, innovative patent that Twitter has willfully infringed upon, you sue them, and they countersue you with other patents that you didn’t even know about that seem obvious and invalid?

The Innovator’s Patent Agreement is a nice sentiment, but the loophole potential is simply too great, and it doesn’t (and can’t) address the fundamental problems and dysfunction in the patent system.

A patented “invention”, even when patented under these terms, is still patented. It’s not free for anyone to use, and willfully infringing upon it is still dangerous and unwise.

Despite constant requests from my customers, I haven’t implemented pull-to-refresh in Instapaper for two reasons:

  1. Twitter patented it, and infringing upon their patent is a serious risk to my business.
  2. Conceptually, it doesn’t quite make sense in Instapaper’s context, so it’s worth neither the risk of infringement nor the potential cost of licensing the patent from Twitter.

Neither of those have changed with the Innovator’s Patent Agreement.

A truly innovative stance would be for a large technology company to avoid filing patents,2 and to lobby aggressively for progressive patent reform to make that a practical choice for every technology company. But that’s not what Twitter has done.

The Innovator’s Patent Agreement is a well-intentioned gesture. But that’s all it is.

  1. If Apple really did implement pull-to-refresh without licensing the patent, Twitter could actually sue them anytime they wanted, unprovoked, because of the second definition of “Defensive Purposes”: Apple has sued other companies for patent infringement in the last 10 years. ↩︎

  2. Many Instapaper ideas might have been patentable, including the one-click-save bookmarklet, many techniques the bookmarklet uses, many text-parser algorithms, some Kindle methods, proportional tilt scrolling, automatic-by-sunset dark mode, smart auto-rotation prompting, smart gesture-error prompting, methods for pagination of arbitrary web content, and methods for inter-app communication.

    I didn’t patent the older inventions because I couldn’t afford to. I probably could have patented some of the newer ones, but I didn’t even look into it enough to do basic prior-art searches. I fundamentally disagree that software patents (and many other types of patents) are a net gain for society, and I can’t participate in that system in good conscience. That’s a stand that I’d like to see more companies adopt.3 ↩︎

  3. Don’t worry, I didn’t patent it. ↩︎

Time and taste

This bad article, nicely rebutted by John Gruber, uses a common argument against Apple: that, inevitably, other hardware manufacturers will figure out why Apple products are so popular, create their own good-enough copies, sell them for much less money, and relegate Apple to the same level of market obscurity that they held with Macs in the 1990s.

People also often apply variants of this theory when guessing how other huge players such as Microsoft, Google, and Amazon will fare when pitting similar products against Apple’s. For instance, Microsoft has effectively infinite money and overwhelming dominance in many software markets. Google has effectively infinite web traffic. Amazon is ruthlessly efficient to sell and ship products at lower prices than nearly anyone.

But all of the money, web traffic, and cheap cardboard boxes in the world can’t buy two huge factors that contribute to Apple’s modern success: time and taste.


Time is twofold: nobody can time-travel to launch a product in the past, and nobody can change how they’ve allocated their time in the past.

No matter how much money Microsoft pours into Windows Phone 7, for example, they can’t travel back in time to 2007 when its limited feature-set and almost nonexistent software library could be more competitive.

And no matter how much Samsung, HTC, Amazon, or Google want to offer high-quality platform software, rich app ecosystems, and well-stocked digital media stores (except Amazon), they can’t change their unfortunate history of minimal investments in these areas over the years.

The iPhone and iPad were built on years of work, experience, relationships, and reputation. There’s a lot more software than hardware in these products. It wasn’t enough to just glue a glass screen to a battery and use an off-the-shelf OS — any hardware manufacturer could have done that. (And indeed, they since have, with some success in phones and little success in tablets.)


Most people don’t have great taste. (And they don’t care, so it doesn’t matter to them.) They usually like tasteful, well-designed products, but often don’t recognize why, or care more about other factors when making buying decisions.

People who naturally recognize tasteful, well-designed products are a small subset of the population. But people who can create them are a much smaller subset.

Taste in product creation overlaps a lot with design: doing it well requires it to be valued, rewarded, and embedded in the company’s culture and upper leadership. If it’s not, great taste can’t guide product decisions, and great designers leave.

No amount of money, and no small amount of time, can buy taste.

(Steve Ballmer.)

Improving poor taste in upper leadership is almost as difficult as treating severe paranoia: people who don’t value taste and design will rarely recognize these shortcomings or seek to improve them. With very few exceptions, companies that put out tasteless, poorly designed products will usually never change course.

Anyone who wants to compete well against Apple is going to need good taste at the top and deep-rooted design values throughout the company.

Might upgrade to the paid version someday

Myke Hurley’s Paying The Price On Android:

Whilst watching a recent episode of All About Android on the TWiT Podcast network, I noticed that as one of the hosts was demoing Plume – a twitter client for his Android tablet – he said that he was using the free version of the app, even though it had ads and that he might upgrade to the paid version as he uses the app a lot.

This isn’t limited to Android — I saw this “plan to buy”/”might buy sometime” sentiment all the time with Instapaper Free for iPhone.1

These users almost never upgraded.

It’s a very common user mindset: they tolerate a lot of limitations, ads, and nags to avoid paying. It’s not that they’re cheap, per se: they just really don’t believe that apps are worth paying for, and they feel cheated or defeated if they end up needing to pay for one.2

It’s not worth trying to appeal to them with a paid app. In most cases, the conversion rate will be so poor that it’s not worth the cost of maintaining two apps and supporting the free users.

App developers can either ignore them as a market — and it’s a big market — or find other ways to pay for their use.

Mobile ads pay very poorly. In my case, ads didn’t even come close to delivering similar value as the $4.99 paid-app sale — I was lucky to get even $1 of value out of an Instapaper Free user. What I’ve heard from other developers and other ad networks suggests that this is pretty close to the industry average.

I decided to yield the free market to my competitors and discontinue Instapaper Free over a year ago, and my sales have remained healthy. (In fact, they’ve increased, but it’s difficult to know whether that was the cause.)

While I’m losing a tiny fraction of sales from people who really would have upgraded from Free to the paid app, I’m also gaining sales from people who would have chosen Free but instead just grumble and accept that they need to pay because there’s no other choice. So far, it appears that I’m coming out ahead financially while reaping huge rewards in simplicity, development, customer satisfaction, and server costs.

This definitely isn’t an Android problem: it’s a user problem. Maybe a significantly larger percentage of Android users insist on free apps than iOS users (it certainly seems that way). But both platforms have much larger demand for free apps than paid apps.

There’s nothing anyone can do about it. Developers just need to decide whether, and how, to address the demand for free software and services without going out of business.

  1. There was never an iPad version of Instapaper Free. And since the iPad’s release, more than half of Instapaper’s sales have come from iPads. ↩︎

  2. Many users also can’t pay, such as residents of countries without methods to pay on iTunes. ↩︎