Why Beautiful Typography is Pixel-Indpendent.

Khoi Vinh a typographer/designer who works for the New York Times has recently delivered some fascinating comments on design for books and magazines and the iPad and the new iPhone. Here is a quote from yesterday's blog:

LESS-THAN-PERFECT VISION

Creating a beautiful display and patting yourself on the back for having good typography is disingenuous, I think. It’s a little like saying a high-definition television set makes for better television shows; an absurd claim at best.

That metaphor is imperfect, of course, because television manufacturers have nothing to do with the content that appears on their devices or with its production. But that, supposedly, is the unique value that Apple claims to offer: they build the whole widget. Not just the hardware and not just the software, but the divine unification of the two into transcendent commercial products.

Steve Jobs’ vision for Apple, repeated in yesterday’s keynote address, posits that the company operates at the intersection between technology and the liberal arts. I think it’s reasonable to regard fine typography as falling within that mandate, but unfortunately, they are falling short of that promise. Building a great display for typography without building great typographic tools is a dereliction of duty. (Khoi Vinh at Subtraction)

There is a lot more in the post and much intelligent opinion in the comments.

While I can understand the typographers frustration at the way that Apple seems to have made some very funny (ie poor) decisions about typography and book design, and one sympathizes with Khoi in his shudder when Apple marketeers expostulate on screen about 'perfect' type, when what we see on screen is a 'typographic calamity'. Yes this is all very odd, and very odd indeed is Apple's decision to embrace the ePUB standard for iBooks and then implement it poorly (that calamitous page would be much better with a ragged right margin). This is odd because the iPad is showing that books really do not need to be as typographically simple and neutered as the ePUB standard encourages or presupposes them to be. But it is important to understand what is going on here. Apple really is just building a platform, they are not, contra Khoi, building the whole widget, or even the whole enchillada, and we notice that Apple are trying to encourage app designers and publishers to work with them. It may be that the most interesting things to do with books that Apple announced at WWDC was its decision to allow iBooks to incorporate PDF files directly (albeit in a different 'shelf'). With the new higher specification screens that Apple is producing the whole motivation for eBooks and the ePUB standard (reflowable text, translation of tables to ASCII, etc) is being undermined. Apple is belatedly recognizing this and acknowledging that many types of books (and most newspapers and magazines) will be very poorly translated to iOS if we need to rely on a file-text format such as ePUB, or even WebKit as a rendering engine. PDF is making a come-back.

The plain fact is that designers and typographers who work for digital media MUST NOT DESIGN their books, magazines, newspapers etc for display on the latest widget. They must design for 'resolution independent' display. Designers should surely notice that we now have 326 ppi Retina Display outputs, but they should then dismiss the fact, retinal displays are here today and gone tomorrow, and designers should be fixing books that will look as good as possible on any of the myriad displays that will be in the market next year, when Apple or one of its competitors will announce 'Leica Display' with 652 ppi. That means treating books and magazines as virtual editions which will be read on scores of different platforms, with many different screen resolutions. PDF files will not be the solution to this challenge, but the fact that they are being pushed back into the user's line of vision is an indication that the way the document looks and hangs together (in abstract, in the clouds, in the database) is still a key consideration. The form in which we virtualise books and documents must capture all the ways in which we might want to read, look at or use them. When we have that right we can deliver the document in as many different resolutions and platforms as we need. When books are virtualised appropriately they do not need to be rendered as typographic calamities.

Why John Battelle really should Like the iPad

http://www.flickr.com/photos/ianus/ / CC BY-NC-SA 2.0

John Battelle wrote one of the best, first, books about Google and I usually read his stuff. But he is underestimating and mireading the iPad over at his Searchblog

....the iPad, just like the iPhone, is designed for vertical integration and distribution lock in. Apple is building its own distribution channel, just as it did with iTunes, and media companies are falling over themselves to make an app for that. Why? Well sure, for once, it's sexy and cool and hip. That's why everyone loved the Wired demo. But the real reason media companies love the iPad is the same reason I don't: It's an old school, locked in distribution channel that doesn't want to play by the new rules of search+social. Sure, you can watch a movie on it. Sure, you can read a book on it. And sure, you can read a publication on it. But if you want to use the web natively, with all the promise that the web brings to media? Not so much. I dont like the iPad because....

About some of this John is right. Apple is trying to build its own distribution channel for books, magazines and movies. As it has already succeeded with music. Of course they are. Furthermore the iPhone does work well for users by both facilitating the web, and standing back from it. The iPhone/iPad has convenient and clever way of cushioning or separating our media experience from the amazing, distracting, open-ness of the web. One of the key points of the app framework which a book or magazine has on the iPhone: it is easier to read the magazine as a seamless and integrated publication. The web may be only a click away, but a properly designed app feels more branded, more integrated, more immersive and more secure for the reader than a set of web pages (tables of contents, archives, page-navigation, credits, layouts, supplements, search etc are supplied within the app cocoon). And I suspect that Battelle is right in that Apple and the big media companies are hoping that the iPad is going to be a relatively tame, safe and biddable media platform: "A sexy version of a portable DVD player-cum-Kindle".

The idea that the app store is destined to be a vast media repository into which old media can simply pour their old wine (and from which Apple draw a fat 30% commission) whilst new consumers charge up flash memory on their iPads, is wrong in one important particular. Apps are different. Books and magazines, at least, are not like tunes. Apps will be much more dynamic and much more transformative of the media landscape than this reassuring picture of the tablet allows. The publishers who get this will make a new vintage for these new bottles.

To see why this is so, we need to focus on two points. Book (or magazine) apps will use the internet in essential ways, even if they are 'self-contained' apps. As we noted in the last post, a local newspaper, is full of valencies to local business and context. A new millenium newspaper should not merely tell you a phone number, the phone number should itself be click-to-call. That is why you need a touch interface on a newspaper now. You do, nobody wants to cut and paste a phone number. Most worthwhile publication apps will be cloud-served, for obvious reasons (currency, communication, search and reference). Book apps will not be like tunes that once synced one simply listens to, or not. They will reference and be referenced. They will engage and they will have to offer services which integrate them with the web. Even if Apple wants to isolate media from the web, no carrier can now do that. You cannot get away from search. No publisher will want to avoid Facebook or Twitter. The web is fundamental. Second, Battelle's picture of the iPad as merely a device for consuming media is fundamentally wrong in ways that he understands. Media on the iPad will be much more enactive, much more intentional (to use one of Battelle's key categories) than traditional media consumption. Media consumption on the iPad will be much more social, more shared and more communicative than traditional media devices. Do you remember the surprise that you had when you first saw youngsters sharing an iPod with two kids each having one bud from the same device? There will be a lot more of that with the iPad -- mostly virtual, mostly over social networks. The intra-device buzz amongst millions of rich media consumer tablets is going to be truly deafening. A book on the iPad will not be merely a book, when it is easy to find out where is the nearest person who is also reading your book, when it is easy to find out what your 'friends' and 'followers' are reading and when it is easy for the book to tell you where you are in relation to where the book is....

Cloud Computing and Content Services

Richard Wallis who blogs as Panlibus for Tallis (the Solihull UK-based library automation specialist which seem to be going places after 40 years of quietly honing their LMS in the black country), has an interesting post on the way that library suppliers are moving into the cloud. Of course that is going to happen, and it will be interesting to see how OCLC, "the 500 lb" gorilla impacts the traditional library automation market. Especially with Google "the 15 million book" gorilla, hobbled by obvious metadata shortcomings, lurking in the background. Following on from Richard's post, I started to wonder how the content aggregators are going to react to the opportunity and challenge of selling content services within a cloud computing framework.

A recent survey paper from Berkeley summarises the hardware innovations of Cloud Computing as follows:

1. The illusion of infinite computing resources available on demand, thereby eliminating the need for Cloud Computing users to plan far ahead for provisioning.
2. The elimination of an up-front commitment by Cloud users, thereby allowing companies to start small and increase hardware resources only when there is an increase in their needs.
3. The ability to pay for use of computing resources on a short-term basis as needed (e.g., processors by the hour and storage by the day) and release them as needed, thereby rewarding conservation by letting machines and storage go when they are no longer useful. Above the Clouds: A Berkeley View of Cloud Computing (p 1)

A service model and a charging system of this kind would be very attractive to content users if it could achieve the radical cost savings typical of cloud computing. Service subscribers would jump at the promise of a service which gives them the 'illusion' of access to infinite information, (this is where the Google library of at least 10 million books comes in) and which eliminates the need for upfront commitments. The first and second capabilities are straightforward, but there does not appear to be a rationale for 3. The problem is that Information is not rivalrous. From the supplier's point of view, whether as creator or as intermediary, there is nothing saved when users do not use a resource. The intrinsic value of copyright resources does not increase because less use is made of them. Paradoxically the value of a scientific resource such as Science Direct actually increases if its usage is widespread. From the information suppliers standpoint the 'trick' is to maintain the illusion that a resource is effectively available wherever it is needed, even though a great deal is being charged for it and the barriers to entry and easy use are high. Paying for the resource on a short-trem or intermittent basis is unlikely to appeal to the rights holder. I suspect that the Books Rights Registry will be slow to sanction Google in the introduction of an hourly 'pay as you go' access model to its main collection.

A solution to this conundrum will emerge, and I suspect that it will evolve in the direction that publishers and rights holders want their information to be accessible, searchable, citeable, and to a limited extent viewable for free. But they do not want to give it all away. The necessity and the attraction of charging for data and content will be limited to services which are in some extent premium, whether by virtue of extreme topicality, of outstanding readable quality, or of additional value services. The Exact Editions service already deploys a cloud-based content management system it will be interesting to see how our partner publishers evolve solutions for end-user pricing and public access (all our public-facing services are now searchable without need for a subscription or controlled access). In fact every page can be viewed at thumbnail size without the need to register an account. Perhaps we need to evolve towards a stage where every citation at least delivers a thumbnail view even of 'closed' pages.

Improving Searchability and Findability

We have enhanced the searchability and findability of the Exact Editions service with a site release which is being rolled out today. The internal name for this enhancement is 'universal preview' and it brings magazines more into line with the way books are handled on the Exact Editions platform.

1. Users can now search every issue of every title in the store. A search for Omdurman will bring up search results and fragments of pages. No full view pages. The first result of a search for Tony Blair is a page which is available in full view (its in a trial issue)
   
2. Users can browse all the content, but only at thumbnail size. If any publisher chooses not to allow archival searching and thumbnail previews, this can be accommodated.
   
3. Contents pages and front covers can be zoomed up to full size.
4. The text of covers and contents pages will get pulled into search engines (Google and others).
5. Existing "full" trial issues are undisturbed, and still feature on the magazine's marketing page.
6. The grid of all issue covers on the marketing pages are now live click-throughs to the thumbnail previews.

We expect that this change will improve the traffic to sample issues, to searches and subscriptions to magazines in the service. There is a cost attached to offering free searching and search results from all the archives (an order of magnitude increase in the content searched from typical queries) but the efficiency gains we have realized since moving to Amazon S3 more than offset the cost of the additional database activity. So this improved accessibility is a direct benefit of cloud computing.

How 2008 Changed the Market for Digital Editions

Keeping the list to manageable lengths, there have been five really important developments this year. Let us work back through the year which as it happens is pretty much the order of importance:

1. October/November: The depression/recession. Hardly anybody saw this coming and, according to one who did, it may be even worse than we have yet grasped. But the depression is a shot in the arm for digital editions. For exactly the same reason that it is now a good time to be in the business of designing wind farms, building solar panels or battery powered vehicles. Terrible news for Detroit, for airlines, for big city newspapers and dire for the advertising budget of consumer magazines: but there is a substantial new business for the magazines, newspapers and book publishers that can deliver efficient digital services. Digital editions tick all the boxes: energy efficiency, global access, low investment, consumer facing, distance learning, transparency, economies of scale, network enhancing etc.
2. October. Google settles with the Author's Guild and AAP. A completely game-shifting agreement. It has still to be finalised and approved by a judge, but it seems likely that the Google system and the new Book Rights Registry will define the shape of the publishing industry for years to come. Digital books will be database-driven, page-oriented, url-guaranteed, access-managed and completely searchable for free. Books will not be freely readable but how they will be sold and subscribed to has yet to be determined. Although the agreement was limited to books in the USA, it seems very likely that the same general regime will apply to books, newspapers and magazines (and other print objects) everywhere. Google Book Search has always been scalable. Game defining and game just started....Publishers of all stripes have still not grasped how much this changes their market, and how big the potential opportunities are for digital publishing in which every publication can be searched, read and referenced through the web. Book publishers are begining to get it, magazine and newspaper publishers are straggling and struggling.
3. July. Apple launches the 3G iPhone in 22 countries. Perhaps as important the first Android phone appeared in October of this year. The iPhone and the Android phones have been designed to be completely web-capable and this takes the digital edition into the pockets of billions of people who will in the next three years be buying mobile phones with instant web connectivity. If books are on the web as digital editions (see point 2 above) billions of people will be able to read them.
4. May. Amazon S3. Amazon's cloud computing was launched in 2006, but this is the year in which it really took off. In May and again in October they reduced charges at all tiers. Cloud computing is the next shape for collective computing and Amazon's cloud computing infrastructure is increasingly attractive to information providers and web services of all kinds. Digital libraries and digital editions are moving into the cloud, and in the same move these buckets of books become shareable and potentially usable by other web applications.
5. January. Amazon's Kindle was out of stock despite being one of the must-have presents for Christmas 2007 (except that it was out of stock in December and remained unavailable until April 08, and is unavailable again until March 09). But since that broadly successful launch the device has gone rapidly sideways. Amazon has not been able to launch the product outside the US. They have been very coy about their sales and their plans for the future of the device. This has definitely not been the year of the Kindle. One could say that it has been sidelined by the iPhone -- why buy a dedicated device to read books when you can read any web-deliverable object through the device already in your pocket? But they have also been scrunched by the Google Book Search proposition -- do we need downloadable books when they are all, always, in the cloud?
   

A Big Cloud Descends Over Europe

And this is not about the recession. This is Amazon's news that they are now putting some of their cloud in Europe. I have to say that I always thought that it always was in Europe. But I guess it is more in the EU now than it was (probably so that businesses that have to say that they host stuff in the EU can stay with that). Even so, it seems to be going against the grain a bit (if clouds have grains). I thought that half the point of cloud computing was that you were never too certain where the computing was. I guess we will move our buckets to the European zone......one of these days (that is the other thing about clouds, no need to be too precise about when they are wherever they are).

Government and Innovation

Why does this Guardian article about the Government putting aside £1 Billion to fund technology startups give me a gloomy feeling?

Somehow one knows that if the government does this, the bureaucracy will kill or at least stifle too many of the innovative companies that go with it. When we started in 2005/6 we wasted too much time (not a great deal thank heavens) talking to and negotiating with VCs and getting a small firms loan guarantee. That was a particularly dispiriting experience because we found out in the end that although our bank pretended to be part of the scheme, it really was not. It turned us down because it had done hardly any such guarantees in the previous two years but remained officially 'part of' the scheme because the government expected the scheme to have the support of the main banks. We managed our start up funding by keeping our commitments very low and at least we do not have to worry about VCs and second or third rounds of lending/funding.

Here is a really good blog by Paul Graham (who is behind the very early stage and light-weight investor Y combinator) on why innovators in information technology now dont need VCs. Here is his argument:

......many Internet startups don't need VC-scale investments anymore. For many startups, VC funding has, in the language of VCs, gone from a must-have to a nice-to-have.

This change happened while no one was looking, and its effects have been largely masked so far. It was during the trough after the Internet Bubble that it became trivially cheap to start a startup, but few realized it because startups were so out of fashion. ......VCs and founders are like two components that used to be bolted together. Around 2000 the bolt was removed. Because the components have so far been subjected to the same forces, they still seem to be joined together, but really one is just resting on the other. A sharp impact would make them fly apart. And the present recession could be that impact.

If the UK Government wants to give a real impetus to IT innovation it should emulate Barak Obama's commitment to 100% broadband availability so that all American kids have access to state-of-the-art broadband. The UK Government could make that happen next year. In schools, colleges, libraries and the streets next year that would make a real difference to thousands of innovators, inventors and startups, a much bigger difference than scores of subsidised VC-style startups, that would only get going in 2010 (if you haven't negotiated with them you have no idea how slow VC's can be, even when they are moving).

Amazon's Cloud gets more Cirrus

Amazon announces its new Content Distribution Network: Cloudfront. This allows their 'pay as you go' web services to be even more widely and efficiently distributed, see a helpful blog at allthingsdistributed. Think of your content as being distributed in the higher reaches of the atmosphere accessible to all, from the cirrus layer. Cirrus clouds being those "which may appear as delicate white filaments, featherlike tufts, or fibrous bands of ice crystals." Cloudnomenclature.

Media will benefit from this distribution of all content through the cloud, perhaps more than we can yet appreciate. When all music, print and video is available in the cloud and available to myriad devices at negligible cost to anyone we will all indulge in this cultural diversity, this trivial availability. Cloud-based access will encourage the already strong trend towards very simple and light-weight electronic and digital devices which can provide access to media.

The Amazon cloud sets admirable standards of transparency and simplicity in its pricing. In their words "Amazon CloudFront passes on the benefits of Amazon’s scale to you. You pay only for the content that you deliver through the network, without minimum commitments or up-front fees." They provide a simple calculator from which you can predict the cost of your requirement [storage, data-in, data-out, put requests etc]. The whole system is so elegant and straightforward that I am inclined to draw unfavourable comparisons with the hugely convoluted, and complex access system and pricing models envisaged for Google Book Search (foreshadowed in the agreement and Registry). Will books really be encumbered by such a bizarre and fragmented pricing and rights management solution? I suspect that the Google proposition may yet be subverted by something much simpler.

Google Book Search Rumours and Rumbles

A week ago there was a report in the Library Journal that the publishers and Google were going to settle their copyright dispute very soon now. No further news on that so far -- and if there was going to be a friendly settlement the Frankfurt Book Fair would have been a good place and a good time to smoke the pipe of peace. My own hunch is that Google will gain more by prolonging this argument than by settling it, so I am not expecting an early compromise. Perhaps there was no subtance to the rumour

The rumble, of elephants crashing around in the forest, is another matter. Something is really happening with HathiTrust, which sees a score of top american university libraries collaborate to produce a giant shared digital library. See an early interview with John Wilkin, its director in the Library Journal. Why elephants? Hathi is the hindu word for elephant, and the name was chosen because elephants remember, elephants are large and they are strong. The focus of the consortium is on preservation and access. On being there when Google has gone. They are already doing some important things which Google Book Search does not do (like being open and informative about what they have). Major university libraries have staying power and I bet this organisation will prosper.

Google Book Search is still the elephant in the library, but the existence of this consortium shows two things. Three years ago major libraries were saying that they could never do the kind of thing which GBS contemplates. Now several of them are collaborating in a much more ambitious project than anybody would have dreamed of in 2002. When we have a universal library in the 'computing cloud' there will be not one, but many literary digital platforms. There will be a whole herd of digital literary elephants kicking around. There will be a lot of platforms to choose from, partly because there is a lot to be done and different ways of doing it. The second and immensely encouraging feature of this new consortium is that it is obviously condoned if not encouraged by Google. The members of the consortium are almost all working with Google and it is to be concluded that Google is keen to see 'collaborator/competitors' in the digital book space that Google has pioneered. Good for Google and good for all of us. The universal library will be open because there will be a herd of elephants. Google may be the dominant male, but not a monopoly .....

Web Services

We are a web service in the strict sense. We deliver a service through the web to subscribers and to publishers. We turn print pages into web pages and organize, search and serve them; and that is all we do. We are a set of pure web services but the funny thing is that we have never hosted our own web services. We never even thought about buying or owning our own servers, let alone running them. Sure we have some physical machines in the office, but nothing critical runs on them. We don't own any hardware that matters. Why buy when you can rent?

From the moment we founded the company and for some years we were happy customers of Vanillamachines (one of the biggest Linux box collocation services) and they gave us a very reliable performance, though nobody from Exact Editions ever set eyes on the 'machines' that we were renting from them. I was never too sure whether the actual boxes were in Texas or the UK but it was vaguely comforting to know that there were some air-conditioned boxes of blades somewhere that were in some sense ours (three or four of them I believe). But a few weeks ago we moved over to the Amazon computing cloud (EC2 -- Amazon Elastic Computing Cloud), and I do not think that anybody outside of Exact Editions will have noticed, and the strange consequence is that our computing is now entirely virtualised. We are using the Amazon computing cloud and there are no physical boxes that anybody could point to and say 'that is where the Exact Editions service resides'. For sure we have a number of 'virtual boxes' in the Amazon cloud, that seems to be the way that Amazon are able to monetize and charge for their commoditised service, but there is no determinate corresponding set of physical chips running our code. For me, the most convincing moment in this transition occurred when our Technical Director noted that, immediately following the transfer, our main database seemed to be running a bit more slowly than he would have liked. He decided to double the size of the virtual server for a day to see if that made any difference. It did. Search and rendering was a lot faster, and as a result he decided to make the upsizing permanent. For an additional $300 a month, (some such relatively modest fee), we had doubled the scale of our database machine and could upsize or downsize them whenever we chose. If we decide that we need so many mega-teraflops for a couple of days, Amazon's computing cloud will allow us to comfortably and temporarily use them. Its this just-in-time flexibility that is so impressive about the Amazon service.

Mind you, perhaps I would have been a bit more nervous about the switch to Amazon's cloud if I had known that they were still in beta. Only today has the beta label been removed from Amazon web services. I guess they did this because they noticed our successful transfer? [feeble joke]

I still occasionally meet publishers who ask if they can host their own Exact Editions service. In future I will have the perfect and honest answer to this request: we don't even host our own service......That is the real beauty of cloud computing. You don't need to host your own. Let the cloud take the strain.

"Low hanging fruit"

Is a phrase I dislike, almost as much as "leaving money on the table". I tend to shudder when these phrases are used because bitter experience tells me that 'low hanging fruit' really only comes to us when we have put in a lot of effort. You need to be able to reach the fruit for it to be low-hanging: make sure that you are tall enough to reach it! Goose bumps from 'money on the table' for a slightly different reason. Consultants use the phrase as though it were a mistake to leave something on the table, but any good deal always leaves some of the benefit (ideally a good chunk of it) with the counter-party. A successful business will always leave lots of money sitting on the table. That way customers are satisfied and they will come back again next year.

These musings are sparked by today's news that on behalf of our publishers we have sold institutional site licenses to the embassy of an Arabic nation in Eastern Europe, a Lycée in Alsace, a middle-sized British charity, the city library of one India's biggest cities (yes India not Indiana!), and one British FE College.

I am certain that modestly priced institutional licenses, delivered via IP addresses which the clients supply to us, are a considerable market. Publishers simply have not twigged this yet. Only in the case of the FE College was there a visible promotional initiative from the publisher. In the case of the embassy there was an alert reaction from our support team -- a matter of confirming that an institutional license, rather than an individual license for the ambassador was what was needed. These institutional sales are mostly happening because the customers are finding that the offer of an institutional license is there for the taking, and they are taking them up because they are there. Of course, Indian libraries will be taking lots of digital subscriptions in ten years time (and I should not be surprised that some are doing so now). But the subs will have to be at a reasonable and affordable price. That means leaving quite lot of money on the ......

I hate to say this, but book and magazine publishers who do not offer site licenses to their digital offerings (who do not have digital offerings so that they can offer site licenses to them) are -- how shall I put this? -- leaving a lot of money on the table, or if you prefer they are ignoring low hanging fruit....It is a funny old world.

Embracing Open Access Publishing

First there was Bloomsbury starting an Open Access academic publishing programme for monographs. Now we see that Springer is buying up BioMed Central, which publishes nearly 200 open access scientific journals. See Peter Suber's comments on this.

I am sure that Springer are not intending to close down BMC (well 95% sure!). I agree with Suber that the really interesting aspect to this acquisition is that the big commercial publishers are now begining to feel their way into a situation in which giving away content, making it freely available, is actually good for the profit-oriented business which also sells subscriptions to some highly prized content. And provides other services. Its what those other services are that is perhaps most unclear at this point, though BMC is clearly doing quite well from its author-generated fees. I am sure that publishers have scarcely begun to think about the ways in which making content freely accessible may be an effective and strange-as-it-may-seem profitable way of publishing. Locking it all up as tight as possible will not work well. Publishers need to go to bed with this mantra buzzing in their brains: the marginal cost of access is approaching ..... zero.

Cloud Computing and Books

We have been blogged before about cloud computing, and thinking about it again. Partly in response to the rather extraordinary views of Richard Stallman reported in the Guardian.

"One reason you should not use web applications to do your computing is that you lose control," he said. "It's just as bad as using a proprietary program. Do your own computing on your own computer with your copy of a freedom-respecting program. If you use a proprietary program or somebody else's web server, you're defenceless. You're putty in the hands of whoever developed that software." Cloud Computing is a Trap.

This obsession with self-sufficiency and self-reliance, veers in the direction of paranoia. You don't necessarily lose control if you outsource a service, especially if there is competition between various service providers. I am sure that there are dangers with a model of cloud computing in which only one company provides a platform for published books (that company would at the moment look like being Google) but there is really no reason why only one company should host and serve print in the cloud. Exact Editions is using a similar approach to the Google Book Search platform and other platforms are emerging (recently launched is Gale's syndication offering, Acquire Content, but there are many more publishers with that potential and several startups trying to become the YouTube for PDF files). There are lots of reasons why we may expect there to be lots of cloud-based book-type services. Amazon itself may migrate from its Kindle-delivered download system to an access-based, lending bookshop in the sky. The desired outcome here is that there should be choice, and continuing innovation. That way we are all safer and we will get better services more quickly.

Stallman shoud be worrying about competition and barriers to entry, not about the inherent collaboration and interdependence that comes from cloud computing, and indeed from the web and the internet infrastructure which is the foundation. At all these levels interdependence is a source of diversity and of strength. So it should be with books in the cloud.

Amazon will it be more Cloud than Kindle?

Its one of life's little ironies that with the Kindle, Amazon has espoused a download method for distributing digital books, approximating to the traditional model of bookshop which feeds individual libraries. In the world according to Kindle we all have our own individual libraries, which are defined by the content held on the Kindles that we own. Google is going with a books as streamed-service approach, in its Google Book Search platform, which approximates to a universal library service. With the Kindle approach there are lots of 'local copies', but with the Google library in the cloud, there may ultimately only be one copy of each title. Google's (and that is what people find worrying).

I find this Amazon position mildly ironic, because one of the few areas in which Amazon is competing directly with Google and possibly winning, is in the provision of Cloud Computing Services to third parties (Google's own cloud infrastructure is much better than anyone else's, but they are not making such a success of engendering third party use of their system). Amazon has an impressive cloud infrastructure of which the S3 service [S3 = Simple Storage Service] is the best known example. Amazon has just announced that it is extending and enhancing its S3 service with an improved method of delivering content rich services. If Amazon decides to switch tack on the Kindle and treat it simply as a blank slate on which users can rent rather than outright buy titles, they will have the infrastructure in place to make this change. Amazon is a true believer in the 'cloud' for next generation computing, but it apparently thinks that digital books are different: droplets on the ground rather than nodes in the cloud network.

How many publishing CEOs know what an API is?

Google yesterday announced their API for the Google Book Search platform (well it had sort of been pre-announced more than 6 months ago) but it has been given more breadth and visibility via this blog announcement. An API or Application Programming Interface is what allows one web service to integrate with and collaborate with another. If you have an API you need to be prepared to welcome the fact that other systems will figure out how to do cool stuff with your data that you never thought possible. The Google Book Search API is pretty tightly constrained, they list 3 API's and I expect that there will be more, but cool stuff will happen.

The system works efficiently and its much faster than some of the clunky preview systems on offer elsewhere. Here is a title from Arcadia that comes across swiftly and informatively. I can well imagine wanting to buy this book after glimpsing it in the Google search and preview mode. But there is a tendency for the Google system to reduce all the titles it manages to a grey uniformity. In some cases, as with this OUP book hosted at Blackwells, the text appears to be literally grey-ed. Is this because many of the titles are being scanned into the system? I am not sure. But I am sure that many publishers who care deeply about the look and feel, the design and readability of their texts, will not be won over to the Google platform.

The Google Book Search API is a great start, but it is not rich and deep. For that Google will need the active participation of publishers, Google itself will need to be committed to a more open Google Book Search, and there will be a lot more work on structuring texts.

Why does it matter whether your CEO knows what an API is? It matters because publishers (and newspaper owners, TV networks, film studios, content makers of all shapes) are not going to allow Google (YouTube, he she or ItTube, or anybody else) to manage and define the API which has access to their content. Having, or buying into, allying with, the API's which manage and accesses your content may be the key decision for media companies in the next decade. Either your CEO knows what an API is, and can find out how, in strategic terms, to negotiate Google's, Amazon's, Facebook's and Apple's, or he/she needs to be a media genius who does it by gut instinct (Rupert Murdoch is the only one of those that I can think of and he is the wrong side of 70). The heads of Random House, Conde Nast, Elsevier, Cengage, Hachette and Pearson really ought to have an intuition about the way their business can develop an API to the servers which are hosting all their content. I wonder if any of them do?

Zoomii

Zoomii is an imaginative way of using and displaying front covers (works fine in Firefox, not in Opera and Safari). It is an alternative interface to Amazon which gives you a good way of shopping for Amazon titles using a 'virtual bookstore' with the covers on shelves, face-out and clickable to purchase or get more data. I really like the way that it is built on Amazon bookstore meta-data, uses Amazon's S3 and Amazon EC2 (Amazon's cloud computing infrastructure) and of course guides you to Amazon's e-commerce system (it should get a promising flow of affiliate income). This is a business built by, with, from, on, and in front of Amazon. Chris Thiessen, the developer has a blog which reads as though it must be pretty much a one man (woman) and a baby effort. Isn't that cute? Isn't the achievement impressive?



One subtlety appeals to me, you can save bookshelves you may have generated. Here is my shopping cart for P G Wodehouse books. Hat tip to PersonaNonData and Brantley.

Reference and Accessibility

As you get older (its more than 30 years since I started out as a greenhorn philosophy editor) you begin to notice that its sometimes quite hard to read the typeface of the books you want to read. Especially when the books are paperback reprints of books that were originally published in a larger hardback format (eg wonderful book on Leonardo). Since the cost of reformatting a major book are trivial, I used to complain about those mean-spirited publishers of great books who did not make any effort to ease the legibility of their republished books for over 50s.

No longer. Publishers are absolutely right not to reformat their popular paperbacks but to leave them exactly as they were in their originally published format, exactly as they were when they were first reviewed. This conclusion was inspired by the kerfuffle surrounding the issue of 'how many ISBNs should a book have?'. See Brantley's posting and reactions summarised on Publishing Frontier. The most extraordinary thing about these discussions is that it appears that many publishers believe that a digital format which does not allow or facilitate consistent citation is an acceptable format for their books to appear in. If the original typography, layout, design and pagination of a book is lost (and all these 'reflowable' formats for ebooks fail in this regard) then it is much harder, perhaps impossible, to devise a consistent way of citing it and referring to it.

When harping on in this respect on the importance of citations and consistent reference schema, within a book and between books, I sometime feel that I may be veering in the direction of millenarianist fanaticism ("prepare for the universal digital library by rendering all pages into consistent web resources, for the digital universe of cloud computing is nigh"). Peter Brantley may even have accused me of such a "born again" approach.

But before dismissing this preference for reliable references, properly evinced by publishers who stick with their original typesetting when they produce a trade paperback in shrunken dimensions, remember that 'Cloud computing' also needs consistent schema for access (so urls matter) and for searching (so proprietary file formats don't help). ISBNs only belong to formats which can be properly cited and searched. Give SKUs to formats which dont literate in the cloud computer. They arent real books so they dont need an ISBN!

The problem of re-sizing pages for the over 50's is going to be solved by our browsers with resolution independent scaleable graphics.

Where Google got the idea.....

Google's Book Search project is possibly their most ambitious undertaking. From one point of view it is an attempt to reverse engineer a proposal entertained by Alan Turing 60 years ago. He was wondering how to design a computer which would have a very large, efficient and affordable digital memory. As a thought experiment he considered the potential for using books ( a library) as a system of machine memory:

We may say that storage on tape and papyrus scrolls is somewhat inaccessible. It takes a considerable time to find a given entry. Memory in book form is a good deal better, and is certainly highly suitable when it is to be read by the human eye. We could even imagine a computing machine that was made to work with a memory based on books. It would not be very easy but would be immensely preferable to this single long tape. Let us for the sake of argument suppose that the difficulties involved in using books as memory were overcome, that is to say that mechanical devices for finding the right book and opening it at the right page, etc. etc. had been developed, imitating the use of human hands and eyes. The information contained in the books would still be rather inaccessible because of the time involved in mechanical motions. One cannot turn a page over very quickly without tearing it, and if one were to do much book transportation, and to do it fast, the energy involved would be very great. Thus if we could move one book every millisecond and each were moved ten metres and weighed 200 grams, and if the kinetic energy were wasted each time, we would consume 1010 watts, about half the country’s power consumption. If we are to have a really fast machine then we must have our information, or at any rate a part of it, in a more accessible form than can be obtained with books. (a lecture to the London Mathematical Society in Feb 1947, quoted by Hodges: Alan Turing -- the Enigma, p 319)

Turing emphasizes the crucial importance of referential transparency in designing book-based machines ("...right book and opening it at the right page, etc. etc...."). There is no point in having a digital book unless the system can locate each and every constituent element within each and every book efficiently. File-based e-book systems have papyrus-like referential opacity. Google Book Search is certainly not making this mistake. Efficient search, random access, referential precision and interoperability will work together in the digital library.

I do not seriously suggest that Google took the idea for their great project from Turing, but it is remarkable that Turing's modest proposal is being captured in ways that he could not have imagined, but of which he would surely have approved.

Windows is Collapsing?

Really? Well two Gartner analysts think that it is. The codebase is too vast and unwieldy. It will not run on small systems. Customers are reluctant to upgrade to Vista. Three or four years ago, this would have seemed like really important news. Now it looks somewhat unsurprising. Exact Editions is a young business (three years) and only typical in the way that canaries are typical in coal mines, but as it happens we scarcely use Windows in our operation.

Our accounts are run through Excel. We word process contracts in Word (but we hanker after a 'click through' alternative). MS Office running on the Mac may be more important to us than Windows. But all the crucial day-to-day software running our platform is web-based (email and web-accessed internal processes). The operating systems that matter to us are Linux first, Mac second, and Windows nearly 'nowhere' -- just ahead of Symbian, which is running our Nokia phones. But all these operating systems are just 'tools'. What overwhelmingly matters to us as we look forward, is the web, and the emergent cloud computer that we all use. Maybe the concept of a monopolistic global operating system has had its day?

The Future of Internet Search?

Mac Funamizu has a very interesting posting which you need to see, as the verbal description is not as compelling as his graphic mockup. Hat tip to Brantley, who emailed me the link.

The first example he gives, of a search on a skyscraper vista, is completely believable. Google has been beavering away with its rich cityscapes. You may have noticed that they added Little Rock, Anchorage and 11 other Metro areas last week. City-scape recognition will be a great way of delivering mobile search.

But I found Mac_Fun's second example even more intriguing. He is right, the device that does city-scape recognition will also do user-generated document scanning. That puts the world of digital documents in a completely different ball-game. The thesaurus example he gives, is a little bit lame......there will be vastly richer linkage patterns available. "Show me who else has read this!" For one. Privacy issues come crowding around mobile search.