Looking over the relative increase in the number of active users for snapmyinfo over the past year led me to wonder, is there a macro trend here? I decided to check Google Trends and came up with this graph:

What this graph represents is the change in Google search volume for the term “qr code” in the United States.It looks like a lot more people are aware of “qr code” in the US and it’s a trend that seems to be accelerating. It will be interesting to see if this trend continues into 2011.

Snapmyinfo’s primary server was hacked

The site was down for about 24 hours starting yesterday January 21 at noon. The main server was compromised by a malicious hacker using a security hole in the Exim program that snapmyinfo uses for email. I brought the site down so that I could isolate the machine and lock it down to avoid any potential further damage.

After studying the attack pattern, I’m certain that no user data was compromised or leaked. The attacker was using a simple “off the shelf” script to try and use snapmyinfo’s email server to relay spam email (even though the script appears to have failed) and was not an attack directed specifically at Snapmyinfo. To be completely safe, I’ve built a whole new server image, re-installed snapmyinfo from source and restored the user database from a backup to be sure that no remnants of the attack remain. I’ve also instituted some new security measures to hopefully avoid this problem in the future.

More people are using Snapmyinfo

Now some happier news: for the past year I’ve been busy working on other projects, but in the interim the number of people using Snapmyinfo has been steadily growing. This includes both the number of signed up users as well as the number of people using snapmyinfo’s email/sms based decoder and free qr vcard generator, qr code generator, and QR code bookmarklet. This may mean that QR codes are finally starting to make inroads with the general public.

Feedback

I’ve put active development of SnapMyInfo on hold for the past year or so because I felt that QR codes were not quite ready for primetime. With the increased activity on the site, it appears that may be changing though. I’ve added a UserVoice powered feedback tab to Snapmyinfo and am really interested in how you think the service could evolve, what features you’d like to see and how it could be improved. I’ve got many projects going on, but feedback could help me determine if the interest in this kind of project is really there and whether it would be valuable to you if there was more development and energy spent here. Thanks in advance!

Generally there are two ways to create QR codes for business cards. The first involves putting the information directly into the barcode. The other usually uses a web address to act as either a landing page filled with the contact information (say your linkedin profile) or as an indirect way of pointing to your contact information. Each method has its pros and cons and I’ll go over each.

The Direct Method

QR codes are designed to store text. Over time, many barcode readers have evolved to understand that when the text is formatted in special ways, that they should do “smart” things with the information. Adding contact information with these formats, in many cases, allows them to be acted upon, say dialing an embedded phone number or possibly rapidly importing a contact into an address book.

One such format is vCard. vCards are fairly common, you may have seen vCards attached to emails with people’s contact information. Unfortunately, there are different versions of the vCard format, so usually only the simplest (first vCard version) is understood by most barcode decoders. While many address book applications can output vCards, just exporting a vCard from your address book application will usually not work and result in confused barcode readers.

If you want to explore QR codes with a simple vCard embedded you can use this form to generate vCard based QR codes. Alternatively, if you just want to convert some text into a QR code, there’s a form for that too.

A typical vCard formatted block of text might look something like this:

BEGIN:VCARD
N:Kowalchik;Michael
ORG:TenZeroLab
TITLE:Founder
EMAIL:mikepk@tenzerolab.com
END:VCARD

Which generates the following QR barcode:

There are a couple of benefits to adding contact information directly to the barcode. The first is that the information can be read without a data connection, all of the information is in the code itself. Another, related, benefit is that there is no reliance on a third party to connect you to your contact information. Most indirect methods rely on an intermediary to direct you to the actual contact information which can present privacy and dependency issues.

One drawback of putting the information inside the barcode, especially if you’re using more complex formats, is that not all barcode readers treat the formats the same way. There’s no official standard for storing contact information, so there’s no guarantee the barcode reader will identify it and do the right (or expected) thing with the data.

Another problem is that if you ever want to change the contact information in the code, you can’t. With the direct method, whatever is in the code at print time is now immutable.

The last drawback of the direct method is that the QR code becomes more complex, “busier”, the more information you store in it. This has direct implications for how easy it is to decode, especially with cell phone cameras. Simpler codes are easier to read. This puts a practical upper limit on what you can store in a QR code printed on something small like a business card.

The Indirect Method

The other method of storing contact information, the one used by SnapMyInfo, is to use a web url to act as a “pointer” to the contact information. This means that the actual contact details are not stored in the code itself, but rather they’re stored online somewhere that can be accessed by the link.

In the case of SnapMyInfo, each user’s contact card is assigned an address. Here’s an example of what’s actually stored inside a SnapMyinfo QR code:

http://snpmy.com/9WDu17f29WQYoe

This link results in a QR code that looks like this (this is the one that stores contact info for SnapMyInfo itself).

These addresses currently turn into an email “mailto” link. If you follow them, your email program should open pointing at SnapMyInfo’s email decoder, send that email and you get contact information in a response. After a lot of trial and error, it turned out that most phone devices seemed to accept maito links and handled the resultant contact data more easily if it was attached to an email. (As an example, redirecting the link to download a contact directly fails on the iPhone. Mobile Safari currently doesn’t understand how to download contact information from the web). One of the nice things about using these kinds of addresses is that their behavior can be altered. A feature I’m planning to add is the ability for users to change what their SnapMyInfo address redirects to on the fly.

Obviously, you don’t have to use SnapMyInfo to use the indirect method. There are other services that are using a similar technique. Also, using a link to something like your LinkedIn profile accomplishes much the same thing, creating a pointer to where you can find your latest information (assuming you keep LinkedIn up to date of course).

The benefits and drawbacks of the indirect method are the inverse of the direct method. Storing a pointer means that you can update and change your contact information over time, and all the barcodes you may have printed remain valid and up-to-date. The indirect method also means that you can store a lot more information “behind” the QR code, the code itself never increases in complexity since the barcode only contains a fixed size web address. This means contact information that comes out can be richer, like the inclusion of contact photos in the contacts that come out of SnapMyInfo.

Using the indirect method, however, also means that the decoder needs to have a network connection to be able to access the data. It also means that the contact information is dependent on the intermediary being there, such as SnapMyInfo.

If you’re interested in giving the indirect method a shot, sign up for a SnapMyInfo account! I would love to hear your feedback.

Click and hold, then drag it to the bookmarks bar on your browser.

Here’s a quick little bookmarklet to generate QR codes. If you click the bookmarklet while you’re on a page, it will generate a barcode based on the address for that page. Try it now! if you click the button, you should get a QR code for this blog post.

Another feature is that you can also generate barcodes from highlighted text on the page. You can use it to quickly encode any page text into a barcode. Of course, if you want to encode arbitrary text, there’s a tool for that as well.

What’s this useful for? Mostly this is just a fun, quick, way to generate QR codes to experiment with. There is one thing I’ve found it’s quite useful for though. Ever wanted to load the page you have open in your browser quickly on your phone? Click the bookmarklet then scan the code using your phone’s barcode reader. It’s faster than typing in long urls! Let me know if you find any other fun or useful hacks!

Update: Had some issues with wordpress and the quote marks in the bookmarklet. Should work now.

Just saw on Techcrunch, that it seems like the QR tests that appeared on Facebook are part of a location service of some kind. At first blush it sounds more like Google Favorite Places, but with Facebook’s focus on people and connections I can’t help but wonder if it’s going to incorporate aspects of Foursquare and Gowalla.

TechCrunch: Those Facebook QR Codes Are Part Of Their Location Plans

A source with knowledge of Facebook’s plans tells us that the QR codes will be used with an upcoming version of Facebook’s mobile app. More specifically, businesses could potentially print out a QR code and put it on a wall or a counter in their venue to allow users to scan it to check-in at that store, we’re told. Facebook is expected to unveil its location plans at its f8 conference in late April.

One of the issues with QR codes that I’ve heard several times is that they’re not that visually appealing. I just ran across an interesting blog post Clothes With Matrix Codes Put Information Up Front | Singularity Hub. The post mentions another technology I wasn’t aware of Data Glyphs, symbols that can be embedded into images in such a way as to be invisible to the human eye:

Speaking of displaying data unobtrusively. Data matrix codes are cool, but they all have a recognizable look about them that is destined to be associated with this decade as much as denim jackets are forever tied to the 1980s. The technology sure to replace them is already here; it’s just proprietary, expensive and impossible to implement by hand: data glyphs. Data glyphs, invented by Rob Tow, who reminisces about them here, are like matrix codes so intricate that they can be seamlessly hidden in unrelated images.

I think linking the physical world with the internet is inevitable. QR codes are an early step and I assumed the next phase would be using facial and other object recognition. It’s also quite possible that these ‘hidden’ explicit links may be the next phase in linking real to virtual.

I also like this quote from the article:

Like telephone poles and printing presses, matrix codes will be remembered in the future as the hallmarks of a fundamentally new and different way of seeing the world. Even if, in the greater scheme of things, they themselves will only be visible for a short time.

Seems like sharing your contact information using QR barcodes has become a popular topic of conversation. Appearing at SXSW on attendee badges and briefly as an apparent prelude to Facebook launching QR codes for profiles, people are abuzz with the idea.

Sharing information using QR codes is a great idea but it’s not without its limitations. I’ve been working on SnapMyInfo for several months now and thought I’d share my experiences with sharing contact information in this way. As I’ve explored the limitations of QR “business cards” I’ve come up with several solutions that I’ve incorporated into SnapMyInfo.

Smart phones and barcode readers not ubiquitous… yet

When I started SnapMyInfo, the original concept was to make QR codes more accessible to more people. The theory is, if more people can decode these symbols then the symbols themselves become more useful. Creating a positive feedback loop in this way could possibly lead to more people using them.

Traditionally the only way to read what’s in a QR barcode is to use a smart phone and a barcode reader. While SnapMyInfo barcodes work with standard barcode readers, SnapMyInfo also has an email based qr decoder that works without installing an application. As long as a phone has a camera and access to email or MMS, they can decode QR codes.

Barcode readers treat data differently

There are lots of different applications for reading QR codes. These symbols have been around for a long time and are an open standard. Over time, these different applications have evolved to read the embedded data in slightly different ways. QR codes store text but if the text is formatted in special ways it signals to some (but not all) barcode readers to take certain actions when the code is scanned (store a contact vcard, open a map, open a browser, etc…).

There’s no official way to store contact information in a QR code. Some common formats include vcard, mecard, and plain text information. If the barcode reader, however, doesn’t understand the format, it may just present you with a screen full of nearly-illegible markup (if it does anything at all).

The one format that nearly all barcode readers seem to honor is that of the humble link. If the text inside the barcode consists of nothing but a hyperlink e.g. http://snapmyinfo.com, the reader application will almost always open a browser to that link (or present the option to do so).

By standardizing all contact sharing using links, SnapMyInfo avoids this uncertainty when sharing contacts. The links used by SnapMyInfo are “smart”. They understand the capabilities of the device that is accessing the link and present the best options for that device. The default on most handsets (for now) is to open up the mail program on the phone to send an email to SnapMyInfo. That email is special in that it will let the system know a connection has been made and reply with the vcard of the scanned user. This means that anyone can scan a snapmyinfo barcode and receive contact information without signing up for the site. Using email for the vcard works across more handsets because most email applications understand attached vcards (even if the mobile browser does not).

SnapMyInfo application links also allow the potential addition of new features to the contact exchange. This includes features like allowing automatic Google Contacts integration for Android handsets as well as allowing users to specify exactly how they want contact share links handled. Dynamically altering the behavior of these links could allow users to create “situational” SnapMyInfo codes that behave differently based on location and time.

Contacts are a two-way street

The organizers of SXSW understood some of the issues I’ve highlighted and decided to use QR code links to link people together. They had some problems, however, because following a link is a “one way” behavior. To address this issue, they created a site you had to log into to complete the “receiver” end of the contact sharing process. This led to a lot of confusion and increased, rather than decreased, the friction for the sharing of contacts.

SnapMyInfo uses email as the way to verify the receiver of the contact for non-logged-in users. Without substantially increasing the friction (just hit “send” on the email) the contact sharing becomes a two way process and doesn’t require any authentication to get in the way of the share. Since the links are dynamic, logged in users can access more features and behaviors, but basic contact sharing doesn’t require either party to be logged into snapmyinfo.com.

Camera phones, not quite blind as a bat

The quality and resolution of the camera on many camera phones (especially so-called “feature” phones) is not great. The barcode reader usually needs a fairly good picture of the symbol to decode these barcodes. Unlike some other barcodes, QR codes can store a large amount of data and grow in size and complexity relative to the amount of data stored.

In addition, there’s also a certain amount of “error correction” that can be added to the barcode. This error correction means that if part of the symbol is damaged, you can still recover the original data. The downside of error correction, however, is that it makes the symbol even bigger and more complex.

The bigger and more complex the QR barcode is, the better picture you need to decode it.

The first way SnapMyInfo deals with this issue is by using links for contact sharing. Something like a vcard, stored directly, results in a fairly large and complex barcode. Even links, however, can result in substantially different levels of complexity in QR codes dues to the variable length and size of links.

http://snapmyinfo.com/blog/facebook-and-the-coming-wave-of-qr-codes/ vs. http://bit.ly/cfITeW

If I encode a link to the snapmyinfo blog and compare it to the same link that’s been shortened by bit.ly, you can see that at the same relative barcode size, the bitly barcode is less complex and has larger features. This makes the barcode on the right easier to decode with less than ideal pictures.

SnapMyInfo does something similar, by internally using a link redirector. You can think of it kind of like a url shortener, except that the thing linked-to can change depending on the device, situation and user settings. This allows SnapMyInfo barcodes to be a standard size and complexity while still linking to a large amount of data. This also allows SnapMyInfo barcodes to contain a higher amount of error correction.

Please wait, “scanning”…

As I’ve tested using QR codes as my “business card” there’s one situation that has come up often as a limitation of using them in this way. At conferences and networking events, if you’re connecting with a lot of people in a short period of time, it can be cumbersome to have them each scan your code in succession, especially if everyone does not have a barcode scanner or some understanding of QR codes already.

SnapMyInfo has added an augmented reality decoder to the set of technologies that can be used to decode these same barcodes. Having a setup like this can enable very high speed connection sharing while giving the people sharing their information immediate visual feedback. This also makes it easier for people to understand what the barcodes represent and hopefully will help popularize their use even more.

It’s not completely obvious what to do with this crazy black and white checkerboard symbol

Probably the biggest issue I’ve come across is the lack of awareness regarding QR codes. This is changing now with big players, like Google, coming on the scene to push the concept. More and more people seem to be recognizing the utility of these physical hyperlinks.

I think with the coming wave of QR codes, this issue should start to resolve itself.

Two dimensional barcodes are in the air. After linking us to Google’s favorite places, storming SXSW, going to the movies with us on Fandango movie tickets, now come the biggest rumblings yet of the coming QR wave: Facebook is rolling out automatic QR barcodes for profiles.

I’ve been working on 2D QR codes for a little over six months now. SnapMyinfo was one of the first projects of TenZeroLab and originally intended only to be a one or two month project. The more I worked with QR codes, however, the more excited about the potential of these barcodes I became.

There’s a convergence of several factors that make me believe we will see more and more of these mobile barcodes. The most important is that smart phones are finally beginning to really take hold in the US. I predict that Android handsets will become the free feature phones that are bundled with wireless plans within the next year.

My original intent with SnapMyInfo was to help popularize QR codes by eliminating the need for smart phones and mobile applications to decode them. The time is rapidly approaching, however, where the distinction of “smart phone” will become meaningless. The majority of people will have a handset that can decode these codes, many with the capability on the phone without having to install an application.

Another factor that’s emerging is the growing awareness of the crossover of information services and physical spaces and goods. We’re on a trajectory that will eventually lead us to an augmented reality future, but for now, being able to explicitly link online information and data in the real world has deep and powerful implications.

I’ve found that SnapMyInfo’s biggest hurdle to date has been explaining QR codes and how to decode and use them. With Facebook’s push to add QR codes to profiles, there is now the potential for millions of new users to be exposed to what’s possible with these two dimensional barcodes. This technology has the potential to be as fundamental as the hyperlink with just as many amazing applications and creative uses.

By far the best way to read QR codes is to use a smartphone with a barcode reader (there are a ton of them for the iPhone in the app store), but what if you don’t have a smartphone?

There are many people who either do not have a smartphone or don’t already have a barcode reader application. When I created SnapMyinfo, one of my original goals was to enable QR code reading for a larger number of people, making these 2d barcodes more useful. I’ve moved snapmyinfo in some new directions lately, but I still maintain the goal of making these barcodes more accessible.

It’s my belief that lowering the technical barrier to using these optical barcodes, even if the decoding method is sub-optimal, will increase their usage, exposure, and adoption. This led me to create a way to read these 2d barcodes without a smartphone. What I’ve built is an email-based QR code decoder that works on photo attachments. Initially this decoder only worked with snapmyinfo-specific QR codes but recently I’ve removed that restriction to start testing general QR codes. The only requirements are that your mobile phone have a camera and be able to send email or photo text messages to emails (e.g. verizon picture messaging). This includes many of the so-called “feature” phones that allow sending text messages to emails.

How do I use it?

1. Take a photo of the QR barcode you want to decode. It’s best to leave some white space around the barcode and try to keep the image sharp. (pull away if you have to)
2. Send the photo to the address shown via email or text message.
3. Wait for the response

This is still test software, so I make no guarantees at the moment, but it should work. The more examples of different devices, their picture, email and text characteristics will help me refine and solidify the decoder.

I’m opening this up so people can test it and let me know their experiences with it. I know I need to improve the formatting for text messages and I’ll work on it as I get example text messages.

I’m still working out kinks in the system so I’d love feedback.

SnapMyInfo now has a simple page to allow you to generate QR codes to experiment with. There seems to be growing interest in these kinds of two dimensional barcodes, so I thought it would be fun to allow you to generate barcodes based on arbitrary text and not just snap codes.

This isn’t the internal SnapMyInfo Barcode generator. While surfacing the internal barcode generator used by SnapMyInfo wouldn’t have been too hard, I often tweak the generator and didn’t want people to have to depend on this external service if I ever decided to modify the barcode output in any way. While poking around the Google Charts API a few months back, I noticed they now have support for generating QR codes. This page is basically a simple front end for generating QR codes using Google’s service. Simply input any text you want (up to 512 characters) and you’ll get a barcode image in return. These barcodes contain text but by using certain text patterns you can tell some barcode readers to do special things with them.

Barcodes as Physical Links

If you create a barcode comprised solely of a URL, including the ‘http://’ like http://snapmyinfo.com/ many barcode readers are designed to interpret the barcode as a link. That means they will, if a web browser is available, automatically take you to the site contained in the barcode.

Telephone numbers

Starting the text with the prefix TEL: instructs the barcode reader that this is a phone number. In many cases a smartphone will dial the number contained in the barcode when it is decoded.

Automatic Email

You can embed an automatic email message in a QR barcode. I’ve found this to be relatively uncommon, but the pattern is SMTP:[email address]:[subject]:[message]. Some barcode readers treat this pattern in a similar way to the ‘mailto’ url protocol in web browsers, creating an email in the smartphone’s email application ready to send.

Example: SMTP:support@snapmyinfo.com:Hello:Greetings to SnapMyInfo

Automatic SMS

SMSTO:7812189522:Greetings to SnapMyInfo!

This one also isn’t as common, but you can create a barcode with the following pattern: SMSTO:[cell number]:[message].

For example, the barcode to the right sends an SMS to TenZeroLab. If your smartphone’s barcode reader understands SMSTO then the barcode will automatically create a text message.

vCard

The barcode can contain a “vcard” as well. vCards are a standard format that’s often used for sharing contact information. This isn’t something you’d typically type directly into a barcode generator, but it does work. A typical vcard in a barcode might look like the following:

BEGIN:VCARD
FN:SnapMyInfo
N:;;;;
ORG:SnapMyInfo
EMAIL;TYPE=INTERNET;TYPE=WORK:support@snapmyinfo.com
END:VCARD

Boston Massachusettes!

Sometimes the vcard standard is also used to embed geographic coordinates in the QR code as well. The pattern looks the same except that a GEO: heading is included with latitude and longitude values.

BEGIN:VCARD
N:;Boston, MA
GEO:-71.059773;42.358431
END:VCARD

Others

There are a few other text patterns that are interpreted by various barcode readers not included in the above list, but these seem to be the most common.

I’m curious how people will experiment with this and what fun and interesting uses people will find.