Thursday, October 18, 2012

And So It Comes to Pass

Back in January, annoyed by the number of people wanting a password lock built into the Kindle - an idea that is frankly naive and problematical - I sat down and wrote what I thought Amazon ought to do, based upon my experience working in security and e-commerce. It became quite a long blog article, which can be found here: "If *I* Was Amazon".

Well, blow me down - they've only gone and done it!

Amazon Whispercast is Amazon's back-end administration tool for organizational users of the Kindle, such as schools, colleges and companies - but it looks as though it would work for families as well. Many of the features will appeal to organizations deploying the new models of Kindles - especially the Kindle Fire HD and Fire HD 8.9" models - such as automatic configuration of wi-fi network connections, VPN configuration, ActiveSync with Exchange servers, etc.

But the basic ability to centralize book-buying, organize users into groups and automatically deploy books to the Kindles is very like the "parental control" requirement. And there's also the ability to create named policies which selectively block access to the Kindle store, block access to social networks, etc.

All in all, it looks like Amazon have done a lot of work on the back end, just as I predicted. I can't wait to check it out, in depth.

Sunday, October 7, 2012

Apps for the Google Nexus 7 and Galaxy Nexus

I'm sometimes asked what are the best apps for the Google Nexus 7 tablet and the Galaxy Nexus phone, so here are a few thoughts, written from the perspective of a grad student or very junior academic. All these apps can be installed from the Google Play store, with no need for rooting or side-loading. They are all free (with the exception of Zandy).

Amazon Kindle
It's hard to escape from Amazon. Whether for leisure reading or for textbooks, Amazon has the biggest selection. The Kindle app is slightly awkward - it can be tricky getting its menu to appear, for example. And a back-lit device isn't the best for reading. But it's nice to have key books available without having to carry the Kindle around as well.

As I've previously commented, the Kindle works best as a reading device - the Kindle infrastructure lets you work with clippings and notes from within a web browser on a PC. I'd say the same for any tablet, let alone the 7" screen of the N7 - set it beside your computer and use a proper keyboard to write.

Blogger

This article was at least partially composed on the N7 while sitting watching TV. The app has support for inserting pictures directly from the device camera as well as from the Gallery or Dropbox. The app can get a little confused if you edit an article from both the web interface and the app at the same time - it starts displaying HTML tags, rather than embedding and interpreting them - but that's a trap that's easily avoided. Great for blogging on the run, with short pieces.

Slices

Slices is a Twitter client with a very polished interface which allows you to organise your Twitter feed into categories (e.g. I have one for Science/Technology).
Slices has a particularly clean design
Zandy

Zandy is an Android client for Zotero. Zotero, of course, is a bibliographic plugin for the Firefox web browser - a highly natural fit, given that a browser is the most common way to access online publication databases. With one click, Zotero can create a bibliography entry from a database like Springerlink, download the PDF document as an attachment and store it in "the cloud".

Zandy uses the Zotero API to access this cloud database. The interface is a little clunky - it would benefit enormously from a rewrite to use the sliding tiles metaphor of Slices and Evernote - but it is easy to forgive this given the convenient access to the database. PDF's can be downloaded, and can be read reasonably clearly on the N7 screen.

Calculators

Disclaimer: I don't like the vast majority of modern scientific calculators - the ones that have an "equals" key - as my brain was molded to prefer Reverse Polish Notation by the classic Hewlett-Packard pocket scientific calculators of the early seventies to mid-eighties.

Free42 is an emulation of the classic Hewlett-Packard HP-42s programmable pocket calculator - it has a very clean and elegant design with a huge selection of scientific functions. It looks slightly over-sized on the N7, and slightly under-sized on the Galaxy Nexus, but works brilliantly on both.

Free42's classic simplicity masks enormous power.

go41c emulates the earlier HP-41C, for those who - like me - are more familiar with its keyboard layout.

For those who need even more power, there's an intriguing option in Frink, a full programming language which seems to incorporate some of the features of high-end graphing calculators, such as calculations and conversions with units.

Evernote

Evernote has a similar "sliding tile" interface to Slices.
A terrific little application for the PC, Mac, iOS, Android - you name it. I use it for the most obvious application - I maintain a work journal with ideas, to-do lists, etc. as well as other bits of writing, so I'm using it as a simple word processor that stores all my notes in "the cloud" and makes them available wherever I am. However, Evernote is a lot more than this: you can record audio and video notes, and you can also use the Evernote Web Clipper to clip web pages (including just the main column) or URL's to notes.

That's just the beginning - there are lots of add-ons and applications that work with Evernote. One that really interests me is the Evernote Smart Notebook - a good old-fashioned paper-and-pencil notebook that works with the handwriting recognition in Evernote, via your smartphone's camera, to record and digitize notes including drawings. Included "Smart Stickers" allow notes to be automatically tagged and organized.

I'd find it really hard to get by without Evernote.

Google Sky Map

Using the GPS, digital compass and accelerometer in the Galaxy Nexus - I presume it works on the N7, too, but it's a bit heavier to hold up - this app displays a map of the part of the sky you're pointing it at. A simple, elegant example of augmented reality in action.
Google Sky Map

Wikipedia

The perfect app for resolving disputes in the pub. Or for looking things up while watching TV.

GTasks

This is yet another app which uses the "sliding tile" metaphor - this time to manage entries in the Google Tasks database. Generally, this is accessed via the Google Calendar web interface, where it sits to the right of the main calendar display, but the Calendar app on Android doesn't show it. All the better for this excellent app, which provides an easy interface to manage your to-do list.

Google Goggles

Another app which makes use of compute power in "the cloud". Point your phone camera at almost anything - a painting, a sculpture, a building, a company logo - and press the button, and Google will look it up for you and provide information.

Google+

Mentioned here simply because it shows up the Facebook application for Android as a clunker. However, Google+ works extremely well with Android devices- take for example, "Party Mode" for Google+ events, which sends all photos directly to the event's page for all invitees to view.

Flashlight

Turns on the phone's LED flash, for short-term use as a flashlight. Trivial, but oh so handy.

Light Flow Lite

The Galaxy Nexus has a multi-colour notification LED, but the phone doesn't make much use of it by default. With this app installed, you get full control: make the LED flash red for SMS's, blue for Facebook notifications and green for emails. You can get just about any colour of the rainbow with this little app.

STVRemote

A bit specialized, but I like it: STVRemote is a remote control for the SageTV HD-300 video extender which feeds our lounge-room TV from the TV server upstairs. I don't think it does anything the standard remote doesn't do, except that since it works via wi-fi, it's not limited to direct line of sight. The same company has a Sage TV client that lets you watch TV on your Android device, but a) it's a little specialized and b) it's also a bit expensive, at over $A28. SageTV is a good news, bad news story: the bad news is that they were taken over by Google and stopped selling their excellent product; the good news is that their technology and ideas are making it into Google TV, so eventually they may produce a worthy replacement.

Do Androids Dream?

Good question. But certainly, not of electric sheep.

I've now had a couple of months with the Google Nexus 7 tablet, and a month or so with the Galaxy Nexus phone, which I bought to match it (you don't want to know how dumb my previous 'smart' phone was). It's been an interesting experience.

While the Nexus 7 hardware represents excellent bang for the buck, the experience hasn't been about the hardware at all. Nor, strangely enough, has it been primarily about the Android Jelly Bean software. It's been about what lies behind it: Google.

In recent years, I've relied heavily on the aging infrastructure of my home office: Lotus Domino servers for email, calendar, web server and various home-grown applications, such as my bibliographic database and PhD work journal, augmented by CentOS boxes running a lot of standard open-source apps for mail, etc. The whole thing was fun to set up, useful for learning and an utter pain in the rear to keep running. Even more nerve-wracking was the fact that my wife's business depended upon it, and any system failures would have devastating consequences.

So I set about a long-term plan to get off it. The first step was to move the email and calendar off Notes/Domino and over to Google Docs. I did this first with my wife's business, and it worked so well that I soon moved my email over. Rather than Notes on the desktop, we switched to the combination of Mozilla Thunderbird with the Lightning calendar plugin with Google provider, and the gContactSync plugin for contact synchronization. The migration proved fortuitous, for a couple of months later a hard drive in our main Domino server died - but by then, it was only running my web site (which I am still working on migrating - another story).

The bibliographic database was replaced with Zotero (http://www.zotero.org) and the work journal shifted to Evernote (http://evernote.com/).

So, it was with this migration achieved that I decided to explore the Nexus 7 (which actually arrived as a birthday gift from my better half). Although I had initially planned to use it as a reading device like the Kindle, as well as maybe to play music, etc. in practice it turns out to be a portable Google box. Initial setup requires a Google account (Gmail or Google apps) but the Nexus happily deals with my three accounts (two businesses, plus university email/calendar). All the calendars are merged but can be turned on and off individually). I quickly installed Evernote and found it to be the perfect complement to my desktop usage (which replicates between my home and work desktops and notebook, the way that Lotus Notes used to).

The lack of wireless broadband (3G/4G/LTE) soon led me to consider upgrading my phone. Kogan (http://www.kogan.com) had the Galaxy Nexus at only $379 - perhaps because the Nexus is under-rated by comparison with the newer Galaxy S III - and since it is the reference platform for Android phones, guaranteed to have the newest version of Android available first, and my existing Telstra contract was nowhere near running out, I decided this was the way to go. The phone came with Ice Cream Sandwich installed, and I left it like that for a few days, but soon unlocked the boot loader and flashed it with the Google factory image of Jelly Bean, so that it matched the Nexus 7. The decision to do this was driven by one key feature of Jelly Bean which makes it a killer: Google Now.

It's hard to describe the impact Google Now has on your time management and phone usage. Here's a little example, which illustrates how it integrates with other Google services:

A company I do some work for wanted me to sit in on a meeting with a prospective client in North Sydney. So, I made sure that my contact there was in my Google Contacts, and scheduled the meeting. Because we would need to go over a few things before the meeting with the client, we'd need to get together somewhere nearby - but where? Google Local provided a list of nearby coffee shops, so I schedule a pre-meeting meeting at one of them. Parking could be a problem, but Google Local sorted that out, too.

I live around 25 minutes drive from North Sydney, and sure enough, about half an hour before the pre-meeting, the Nexus 7 chimed that it was time to leave. I've previously noticed from experiments that Google Maps and Navigate have good information on traffic density, and will pick the best route - it can offer me at least two different routes to work, for example. So I clipped the Galaxy Nexus into a windscreen mount and let it navigate me to the car park, and once out on the street, I used it to find the coffee shop. Shortly before the client meeting, it alerted us in time to walk there, and I used Evernote on the N7 to take conference notes - which I then emailed to my client.

It's the ability of Google Now to present information before you need it that is so radical. There's an article in MIT Technology Review that stresses this point (http://www.technologyreview.com/news/429345/googles-answer-to-siri-thinks-ahead/) as well as an interesting article on Google's Knowedge Graph (http://www.technologyreview.com/news/429442/google-puts-its-virtual-brain-technology-to-work/).

The Knowledge Graph lies behind some of Jelly Bean's other neat features. For example, while using it for navigation, I'd noticed that the voice prompts include information read from road signs - this is obtained from the Google Street View camera cars and integrated into Google's maps databases.

The other really neat Jelly Bean feature that depends upon the Knowledge Graph is voice search:


I haven't done a direct comparison with Siri on the iPhone, but I suspect she wouldn't come out of it too well.

Apart from these killer features, there are lots of other things to like about the Galaxy Nexus / Nexus 7 combination: transparent synchronization of all contacts, email, etc. It's not without problems - for example, each of my Google accounts contains my wife as a contact, with various different email addresses (because we're in each others' businesses and they're both now Google Apps domains), but Google also wants Google+ to become a central feature of the Google world - so my better half has wound up with multiple entries in my consolidated "People" app, and Gmail invariably picks the wrong email address when I start to type her name. However, it's rather cool to see someone's face, picked up from Google+, when they ring you.

I'm learning to love Google+, and I'm coming to agree with those who say that it's been a sleeper that is now rapidly growing to rival Facebook. I have very different circles on G+ compared to groups on FB, so there's no rivalry at this point. However, there's no doubt that the tight integration of Google+ with the other parts of the Google ecosystem and especially with the Android apps makes it far more functional than Facebook. For example, scheduling an event through Google+ puts it straight into the calendar, and on arrival at the event, it prompts to put the phone's camera into "party mode" in which every photo is automatically uploaded to the Google+ event page for sharing with the invitees.

Similarly, there's tight integration with Blogger/Blogspot, Youtube, etc. - all of which are supported on the Android devices. This does lead to one of the other sources of pain: getting your Gmail accounts sorted out. Both my wife and I had ordinary Gmail accounts, started before we'd migrated our business email to Google Apps. I'd started this blog using my old Gmail account, and she'd started her blog using her old Gmail account. Now, while I was easily able to add my new Gmail account to this blog and then take over administration - I want to do everything while signed in as me, myself and I, just one account - for some reason she ran into terrible trouble doing the same process. At one point, we thought she'd completely lost access to her blog - months of blog entries effectively lost - and we had to create yet another Gmail account for her as part of the transition. It was a tricky, nerve-wracking process, which leads to another major recomemendation:

Before you buy a Nexus device - phone or tablet - make sure that you get your Google accounts consolidated, rationalized and squared away, with everything running smoothly with one primary account you intend to use. Multiple accounts are certainly possible and useful - as I say, I have two business accounts and one university account - but you really want to have as much as possible consolidated to one primary account.

I do use the Nexus 7 for other things: I have the Kindle app for various books, especially textbooks, and I use Free42 and go41c as replacements for my trusty - but aging - Hewlett-Packard programmable calculators. I dabble on Twitter using Slices, and I shop on eBay using the official app.  I even use the Galaxy Nexus to talk to people, occasionally, but I have probably taken more photos than made calls, to date.

However, for most applications, it seems that most of my data - certainly, most of the data that organizes my life - has moved into the cloud. I access it via a conventional computer, or I access it via the tablet, or I access it via the phone - but in most cases, it seems that Google has it, and places it before me before I need it.

So, to answer the question: No, I don't think Androids dream - in fact, I'm not sure Androids even sleep.

Sunday, August 26, 2012

Vale Neil Armstrong

"I am, and ever will be, a white-socks, pocket-protector, nerdy engineer -- born under the second law of thermodynamics, steeped in the steam tables, in love with free-body diagrams, transformed by Laplace, and propelled by compressible flow." - Neil Armstrong

One of the reasons I personally take umbrage at the decline of the United States as a leader in science, engineering and just plain rational thought is that I am a child of the sixties, and like many of my generation I remember staying up late to watch those blurry black-and-white images as he descended the ladder and set foot on the moon; at that time America seemed to be a beacon, showing the world the way forward to a future in which science and technology would solve all our problems.

It didn't come to pass. Perhaps we were naive to think that way.

However, the moon landing inspired me to pursue a career in science and engineering, albeit in a very minor way. And eventually I learned to fly, too.

To me, the death of Neil Armstrong represents a punctuation mark in American history. He was a pilot and an engineer, what some call a "maker" - one of those who don't get distracted but knuckle down to solve problems, important problems, who put in long hours to master complex skills and who put their lives on the line, both in combat and in test flying. That kind of life and that set of values, I suspect, give you a very different view of what's important and what's not.

These days, our would-be-makers are giving us toys and straining to write an app as successful as Angry Birds (which I freely confess, I have never played). It all seems rather trivial, and that consumer technology has overtaken aerospace is demonstrated by NASA's plans to use Android smartphones as the computing power on a series of small satellites. Armstrong, Aldrin and Scott made it to the moon and back with less computing power than most people carry around in their pockets today.

Meanwhile, the rising tide of anti-intellectualism in the US brings us prospective leaders who don't understand enough basic science or even basic rationality to have any hope of addressing the key issues the world - or even just the US - faces today, such as anthropogenic global warming, public health issues, the biology and ethics of reproductive rights and a lot more.

I can only hope that some of those hoping to "lead the free world" will remember the emotions they felt when Neil Armstrong showed how America really led the world, and then realise that their current approach isn't working. But I expect to be disappointed.

Sunday, August 12, 2012

Googling My Life Away

I'm a big advocate for privacy, but I've recently begun to understand the temptation to trade it for convenience in a digital lifestyle. My family gave me a Google Nexus 7 tablet for my birthday, and I've been seduced by its convenience.

I'd been using Google Apps for a while anyway, as part of a strategy to switch away from my own mail and calendar servers, which were becoming increasingly expensive and time-consuming to maintain; the Nexus tablet seemed to offer easy integration with that, as well as the University email, etc. However, within a couple of days, I'd become an app junkie, having installed Evernote, the Kindle app and Drop box to give me access to my work files over coffee. Next came the Wikipedia app, followed by emulations of my favourite HP calculators.

What impressed me most, though, were the mapping applications integrated into the system. Sitting at home, I can hit the microphone icon in the Google search bar and say "Navigate to Macquarie University", and within seconds, I've got comprehensive directions. If I set out following them, the Nexus 7 will provide voice prompts and a continuously-updated display, only getting into trouble if I deviate from the suggested route - being wi-fi-only, it is unable to update its maps and speech prompts, although it still seems to provide correct left/right arrow prompts on-screen. And of course, its "Local" app can provide listings for nearby caf├ęs, restaurants, pubs, ATM's, etc. - all very convenient. But to do this, Google needs to know where I am. Should I worry?

To analyse this, we have to tease apart several notions which contribute to, and partially compose, the concept of "privacy". First, there's essence of Google. I tend to the idea that my privacy is compromised when another person knows something about me which I do not want them to know. In this sense, Google is not a person - it's just a cloud of servers which are continually processing data on my behalf - email, appointments, my web searches and now my whereabouts. No person is watching, and my privacy - so far - is no more compromised than it is when anyone glances in my direction as I pass them in the street.

Of course, if someone wants to track my movements, there's now a central point where they can look. Google - the corporation - obviously is interested in directing targeted advertisements towards me; I've never been particularly concerned about that for two reasons. One, I rarely see advertisements anyway, thanks to AdBlock Plus (and I never paid much attention to them anyway when I did), and two, I used to work in the direct marketing and advertising business and actually see the benefit of targeted ads that are more likely to be of interest, compared to the shotgun blast of scattered general media advertising.

The more important question is whether anyone else can get access to 'my' 'private' data. At this point, the obvious candidates are government agencies; Google being a US corporation, it is most likely to open my kimono in response to requests under US laws like the PATRIOT Act. But then, I've always assumed that everything I send over the Internet is intercepted and processed in various ways anyway, and it's never bothered me. I do use encryption as routine - I use SSH when logging into my own servers, for example, and a VPN tunnel from my laptop back to the office - but that's to secure against independent hackers and not for privacy.

Coming back to the notion of privacy as an unauthorised disclosure of information about one person to another person - in this case, the information being disclosed is my location. To me, that's not particularly sensitive; most of the time, I'm exactly where anyone who knows me would expect me to be: sitting at one of my desks in front of a computer. If I ever do want to cloak myself in secrecy, I can just turn off GPS or turn the device off altogether.

The other seductive aspect of the Nexus 7 (and its Galaxy Nexus phone cousin) is the ability to tie together various accounts and services, such as Gmail, Google+, Google contacts, etc. I wrote the bulk of this article on the Nexus tablet, and when I post it, it will quite possibly be linked to my Google+ account, might well have my location attached, and so on. But this isn't an issue of privacy - I'm publishing the article, making it public, and I have sufficient control over what's disclosed.

So, all in all, I'm surrendering to the temptation and letting Google into my life. I don't think I'm in danger of being reduced to a carefully-monitored profile, but if I do, I'll let the world know about it.

Monday, January 2, 2012

If *I* Was Amazon

There has been a lot of requests from one or two people for "parental controls" on the Kindle e-reader, and even more for controls on the Kindle Fire. The problem is that different people have different concerns: for some, it's avoiding accidental (or unauthorized) purchases of any kind, for others it's blocking access to erotica or other genres and still others are have more general concerns about all kinds of content available to the device's web browser.

A common suggestion for dealing with unauthorized purchases is the application of yet another password for authorization in a pre-purchase state. I see two problems with this: first, Amazon's business model revolves around making purchases easy - they even have a patent on one-click purchasing, so there are clear commercial arguments against it. Secondly, the questions on the Amazon Kindle customer forum indicate that people often forget passwords and get confused over which password to use when (a point I've already addressed in the context of wi-fi configuration).

Blocking access to the Kindle Store on a per-genre basis is also fraught with problems. First is the fact that, in order to increase their visibility to potential purchasers, publishers will list their books in as many genres as possible. It's entirely likely that, in the face of widespread filtering, erotica publishers would move many of their titles - especially free samples - into "romance" or similar genre classifications. In response, users might start to block multiple genres and before long, we'll have an unwieldy interface and more load placed on Amazon's servers (and slower response times, too).

If I was Amazon (and I'm not, but I do have some insights, since I teach a course on e-commerce technologies) my top priority would be allowing multiple sub-accounts. A sub-account would represent one person and their reading devices, while the master account would be used for billing and administrative purposes (adding/deleting sub-accounts, authorizing/disabling purchasing by sub-accounts, etc.). This would not require any modification to the Kindle firmware - all the development would be at the server end - but it's non-trivial.

The number one goal for this would be to attract education and corporate accounts, e.g. teacher/librarian could purchase books and then distribute to students' Kindles, etc. As a revenue-generating enhancement to Amazon's product offerings, this is far more likely to get done.

Now, this would require negotiation with the publishers as it would break the current model where one account = one reader = one sale. It would open the possibility of a new bulk-purchasing or -renting model for institutional accounts, with corresponding pricing. However, for small group use - e.g. sharing within the family - it's not a huge extension to the notion that a single book can be on six devices; it just means the six devices could be six different people within one family. Nonetheless, it does have legal/marketing ramifications which would require a) rewriting of existing T&C/licenses, and b) a small leap of faith on the part of publishers.

After these enhancements have been introduced, Amazon would have a flexible platform which would meet many of the needs of concerned parents, who could then selectively purchase books for their children's sub-accounts. It would also meet the needs of couples who often want to share their books and have long done so with paper books.

However, I don't see any kind of "rating"-based or genre classification-based parental controls system as being practicable, at least with current technology.

More Detail

 The way I see it, you would allocate devices to accounts - just like we do now - but with the definite understanding that an account represents a person (after all, that person can log in with their email address and password to manage their account, within the limits set). Then we'd allocate the accounts to the "master" account, which is also associated with a payment device (i.e. credit card).

Then the master account could do things like creating collections and granting the sub-accounts access to each collection, enabling sub-accounts to buy or not, etc. Creating collections would be more versatile and easier than allocating individual books to sub-accounts, although I think that should be possible, too. The archive for each sub-account/person would contain just those books in those collections to which they've been granted access.

Here is what we software architects call a use case, in a parental controls context: Mom creates the master account, adds credit card information and then creates sub-accounts for Alice, Bob and Carol. Since these are sub-accounts and have no credit card, they have purchasing disabled by default. Then she creates a "YA" collection and buys a few carefully-selected books which she puts in there, and then grants A,B, and C access to the "YA" collection. They can choose to download those if they want, and they probably will. Carol is 18, mature and trustworthy, and so Mom might grant Carol the right to buy books and add them to the YA collection, too, using the master account credit card (Mom always gets the "Your Amazon.com purchase" email so she knows what's going on").

It would be beneficial if Mom didn't have to create the sub-accounts, but could simply nominate an existing account as a sub-account. This would make upgrades easier, particularly since some families (like ours) already have separate accounts for each person. In this case, the linked "child" account could perhaps have an option to retain ownership of its own books, or to transfer some or all of them to the "parent" account. By doing this, I could create a "family" account, then my better half and I could link our accounts under it, transfer those Peter Robinson "Inspector Banks" novels we've independently bought to the "family" account, and share like we did with all the paperbacks for many years. (I defy any publisher to tell me this would be an unreasonable thing to be able to do!)

And of course, Mom and Dad's "Adult Fiction" collection would be off-limits to the kids - only Carol has even thought what that might be and doesn't *ever* want to conjure up *those* images again. (I'd name that collection "Parenting Manuals" anyway.) Visibility of collections to which access is not granted is an open issue - if Alice and Bob want to know why they can't have access to the "Adult Fiction" collection, Mom (or Dad) should have an explanation necessary. I rather like that. If the kids are studying "Human Reproduction" at school and want a book about it, then Mom (or Dad, but I bet it's Mom) will have to be involved in selecting an appropriate book and putting it in the correct collection.

I would would allow a "master account" to configure the one-click purchase option on the device/sub-account to operate in one of three ways:
  • buy (in other words, the current operation)
  • do nothing (one-click purchasing disabled, for use in schools and institutions)
  • add product to a wish-list
Now everyone wins. Parents could disable one-click buying altogether - the kid either doesn't get the option or can click uselessly on the button. Or they can let little Bob and Carol click to let Mom & Dad know they really want something - Mom can review and approve (i.e. "gift") the purchase, or use the wish-list to choose birthday and Christmas presents. Amazon is happy because kids can use the wish-list to exert "pester power", increasing Amazon's sales. And the existing status quo is maintained for master accounts or perhaps disabled by default on sub-accounts.

Amazon already has the code to add things to wish-lists with a single click - their internal systems are based on what IT people call a "service-oriented architecture" anyway, so this wouldn't be too expensive to set up. Way better than the "stick a password on it" approach. 

For the classroom, the collections idea could be used to group all the books to be used by a particular class, or to group all the books suitable for a particular age group, or related to a particular topic, etc. An obvious related convenience would be the ability to create "classes", i.e. groups of sub-accounts, so that all the kids in 3B could be referred to as "3B" and then "3B" is granted access to collection "3B" plus collection "Plants & Animals" and so on.

Obviously, the sub-accounts in a school would not be able to buy books, but they could certainly download the books from the archive collections to which they have been granted access. The teacher or school librarian would have buying authority for the master account.

This would make Kindle use in the classroom significantly more attractive to schools. At the moment, a few brave teachers are literally collecting all the Kindles, taking them home, and then one-by-one turning on wi-fi, downloading the right books, turning off wi-fi and then taking the Kindles back to school the following day - an onerous task that I'd consider above and beyond the call of duty. With the scheme I'm outlining, the Kindles could remain in the school, with all management being done via the web interface for the master account. The kids themselves would connect via wi-fi and download the books they need.

By the way, I've used the term "sub-account" but really this is just the existing Amazon account, with some extensions. What I'm describing is two things, from a technical perspective - the establishment of a parent-child relationship between the existing Amazon account structures, plus the addition of some access control functionality, with the ability of the "parent" account to set the access controls or permissions for the child account. Accounts could even have many-to-many parent-child relationships - for example, an individual's account might be a child of a "Family" account for household sharing, but also a child of a "Book Club" account which allows the members to share access to a book they're reading and discussing. In essence, this is a generalized architecture to manage sharing of digital content with controls to meet the needs of both publisher and customer, in a variety of situations.

Licensing

Which raises the issue of licensing. For family use, I could make a pretty convincing argument that it would simply legitimize and make manageable what people have been doing for years by putting multiple Kindles on one account and sharing books under the "six devices" rule, and hence would have minimal impact on publishers' revenues, while creating goodwill. In some jurisdictions, such as Australia, copyright law already recognizes the right to make copies of some digital content for personal use within the household. However, I can't see any way of getting this done without publisher involvement - remember, we customers don't buy ebooks, we buy a licence to read them and the current license does not provide for what I'm describing. And the same mechanism should apply to all other forms of digital content - I've been describing this in terms of books, but it can and should apply to movies, music and apps.

Licensing should ideally be handled on a per-device basis, like now, with the notion of a "site license" as currently used for software. A family could probably be accommodated within the existing "six-device" limitation we have today - a large family might have 8 sub-accounts, but 6 of those are the kids, and so there wouldn't be more than 6 copies of the "YA" collection books in use at one time (after Mom has returned her copy to the archive). Effectively, the publisher really has to agree to the terms of the licence and set a realistic price - from that point on the technicalities of enforcing the licence is all down to Amazon, so the publishers needn't get involved.

Obviously, a "site license" for a school would require enough simultaneous device-copies to accommodate an entire classroom plus teachers, etc. although not all copies of all books need be in use at any one time. Amazon already has a book lending mechanism which can be used to remove books from a Kindle after the loan period has expired, and this could be used within an institutional environment to provide library-style functionality, with a certain number of books "loaned out" at any one time and automatic "recall" if a new borrower requires a book after the loan period has expired.

If I Was a Publisher

If publishers don't get to grip with these kinds of requirements, they're going to starve. I believe specialist textbooks publishers can already do it, although I don't know what their terms are. Ultimately, though, the economics of market forces are going to to corral the publishers into the right direction, because if the publishers don't respond, this will create an opportunity for new agents and publishers to move in and provide representation and marketing services to existing trad-pubbed authors, in part by getting to grips with these challenges and opportunities. Competition will then force the remaining traditional publishers to break ranks and - there's no better way to say it - get real.

If the publishers fear the market dominance of Amazon, then they should seize the initiative in doing what I described above: offer innovative licenses which allow sharing and within-account lending, whether for institutions (big $$$), families (smaller $$$) or whatever. And then let online retailers - whether new entrants or established players like B&N, Amazon, etc. - compete to provide the best implementation, service and functionality within the terms of those licenses.