In this Episode
Bill Kasdorf is a partner at the consultancy Publishing Technology Partners, where he provides guidance and insight regarding the publishing industry. He has worked with many well-known companies and institutions, from the scholarly press at Harvard, to the World Bank and the British Library, amongst many others.
In this interview, Leanpub co-founder Len Epp talks with Bill about his background and how he got interested in the world of books, the importance of standards and interoperability when it comes to accessibility, how standardization fuels innovation, changes to the publishing industry in the past decades, from scholarly to trade publishing, and how the EPUB format is managed, and other topics.
This interview was recorded on September 18, 2018.
This interview has been edited for conciseness and clarity.
Len: Hi, this is Len Epp from Leanpub, and in this episode of the Backmatter podcast, I'll be interviewing Bill Kasdorf.
Bill is a partner with the consultancy Publishing Technology Partners, where he and his colleagues provide guidance to companies and service providers in the publishing industry. Bill has particular expertise in standards alignment, accessibility, and a number of other important areas.
Bill has received a number of industry awards, including the Distinguished Service Award from the Society for Scholarly Publishing. He's also on the steering committee of The W3C Publishing Business Group, and Chair of the Book Industry Study Group's content structure committee. Bill's also Principal of the publishing consultancy Kasdorf and Associates, and his client list includes some names I'm sure our listeners will have heard of, including the Harvard and MIT Scholarly Presses, and organizations across a wide range of activities - including the World Bank, the EU, and the British Library.
So, thank you, Bill, for being on the Leanpub Backmatter Podcast.
Bill: Glad to be here. Thanks for the nice intro.
Len: Thanks. I always like to start these interviews by asking people for their origin stories, and I was wondering if you could tell us a little bit about where you grew up, and your journey to how you found yourself first involved in book publishing?
Bill: Well, I sure can tell you about that. I grew up in Wisconsin. You don't want to hear my real origin story, because it involves how I met my wife, and how I determined my entire career for the rest of my life.
I got into graphics - I was an English major in college, so probably the relevant origin story is that I was very active in the student union at the University of Wisconsin, which is a really active, unusual group. In fact, I'm still a trustee of that organization 50 years later.
I chaired the literary committee, which brought in the visiting poets and writers, published poetry chapbooks, did a literary magazine, etc. So to do that, I took a typography course, so I could do handset type and do cool books. I really just got into both publishing, but also the technology of publishing. It was like, how is this actually done? So that's my origin story.
I've basically been in the production side and the vendor’s side - editing, design, typography, etc. I was also a really interested in generalized markup schemes, because when I started out, all the tools were proprietary, nothing worked with anything else. I just thought, "That's crazy. There's got to be a better way to do this." So I was really, really early on in that. And that led to SGML, that led to XML, that led to EPUB, that leads to web publishing - which is what we're working on right now. So that's how I got to where I am today.
Len: And for those listening who might not be aware, what do you do when you do handset type?
Bill: It's actual little chunks of lead that you put in a little - I can't remember what you call it now. But anyway, it's a little tool where you just line them up, and then you put them in line by line, in a framework called a chase. And then you ink it up, and you put it in a press - well, you put in the press and then you ink it up. Then it presses paper against the type, and there you go. It's called a letter press, because actually pressed into the paper.
Len: I'm really curious, how long would it take, for example, to do the front page of a standard newspaper?
Bill: First of all, that's done by many people at the same time. So it's not one person doing that. But if it were one person, wow. Hours, and hours, and hours. That's a lot of little letters. I mean, we're talking about 17th, 18th century if we're talking about handset. At the time when they were done from metal, they were done with casting machines called the linotype or the monotype machine, where operators are sitting at a keyboard, and it's actually casting a whole line worth of type at a time. So they're not making it letter by letter.
Len: You mentioned you had an early interest in markup schemes. I was wondering if you could talk a little bit about what those are, for people listening who might not know?
Bill: That's a, that's a really good question. I actually do a lot of speaking and writing on this subject, and I always stress that people today, for example, get really obsessed about XML. I'm known as an evangelist for XML, and I still am, although there are certain technologies are moving beyond that now.
But fundamentally, it's about being able to identify what the constituent parts of your content are, and then tagging them with consistent tags that basically say, "This is a heading," "This is a heading level one versus heading level two." So, now I'm kind of implying some structure in my content. "This is a quote," "This is a footnote," etc.
Back in the day, you just made those things up. And in fact you still do, people doing a style sheet in InDesign typically just make up the names of what those styles are of those things. But that's really unproductive, because then nobody else knows what they mean.
So when I talk about a markup scheme, I'm really talking about something that's a standard that enables interoperability. So, if I use HTML to markup those things, there are millions of people that know what they are. And there are millions of systems that can work with that content, so that the computer knows that that's a heading, and it knows that that's a new section that's beginning - and knows that this is a quote, knows that this is link, etc. And of course, that plays right into accessibility.
That's why the evolution of these technologies getting more and more interoperable is what ultimately is making content. Where we're going with this, is what we call "born accessible content". Where it used to be that to create an accessible file for a person who's print disabled, for example - like a braille reader, or a screen reader, or something like that; well I guess predating screen readers - you had to use very specialized markup that their tools were programmed to interpret. And most publishers not only didn't bother, they didn't even know how to do that stuff, and just didn't do it. So, very little content was accessible. Now, the same technologies that make content accessible to somebody who needs assistive technology, are the technologies we're using anyway - like EPUB and HTML.
Len: I've got a lot of questions to ask you about accessibility a little later, but before we do that - I think probably everybody listening nowadays is familiar with what HTML is - it's something we use to write websites. But what's XML?
Bill: XML's actually a predecessor of HTML. HTML is a vocabulary and a syntax, right? In other words, HTML says, "Here's the tag for a section. Here's the tag for an aside. Here's the tag for heading level two." "It's in H2." It's expressed in angle brackets, and uses a particular kind of syntax. XML is a level more abstract than that, because it's not the vocabulary - it’s the way of encoding a vocabulary.
Scholarly publishing. for example, uses an XML model called JATS - Journal Article Tag Suite, or BITS - Book Interchange Tag Suite. That's a whole different set of tags, but they both use the XML - it's like a meta-language. It's a way of expressing a tagging scheme. So Docbook is an XML scheme, DITA is another XML scheme that's used in technical documentation. There's all kinds of different vocabularies that are written for particular purposes, and they can all be expressed as XML.
What that means is that they're machine processable, because you not only have the tagged content, but you've got what's called a schema or a DTD - a Document Type Definition, that formally documents, "Here's what these tags mean and here's how they relate to each other. And here's what you can use here and what you can't use here." Your DTD, your schema could say - it'll beep back at you in effect, if you've got an H1 followed by an H3. And it'll say, "No, no, no, no. You can't do that. You have to have an H2 in there next."
Len: And what were the forces that brought about XML? When was it created?
Bill: It was created, actually, from SGML. And actually, when I said that it predates HTML - I don't think that's actually correct, now that I think about it. SGML is a Standard Generalized Markup Language. That came out of work done largely at IBM, called the Generalized Markup Language - GML - way back in the day. This goes back to the early 80's - late 70's, early 80's. And there was work done in the Association of American Publishers - AAP DTD. There was a DTD in scholarly publishing called the Majeure DTD [?]. This is back in the SGML era.
But SGML got really complicated because, frankly it was too flexible. Remember I told you that with HTML you've got these little angle brackets that delimit the tags?
Bill: With SGML you can use whatever you want as those delimiters. You just have to say what they are. So now, guess what? You've got a mess on your hands if those files are out there and you expect systems to be able to use them. So XML came along as a simplification of SGML. It said, "No, no, it has to be angle brackets. You have to nest everything properly."
I mentioned that you have to have an H2 after an H1 and an H3 after an H2, that's called nesting. This one section is nested inside another. That's called well-formedness in XML. If your content doesn't do that, it's actually not XML.
The other thing is that it uses an encoding scheme for the characters called Unicode. So an A is universally known as, "blah, blah, blah" value in Unicode. So whatever language, or whatever typeface you're using or whatever - in XML you already always know when you see that when you see that sequence, it's an A, or it's a B, or it's a C.
Len: It's literally a sequence of zeros and ones that's standardized.
Bill: Absolutely right. It's all binary. Unicode really encompasses almost all, not all, but almost all of the world's languages and scripts. Huge numbers of characters are in Unicode.
Len: I have a question about your story. I didn't know that you majored in English. But from English major to English major, who ended up in a technical area, I'd like to ask you how that happened for you?
Bill: Well, I think going back to my being an English major at the university, and being so involved in this literary group - and being involved in publishing things as a part of that.
But I'll tell you an interesting anecdote that your question prompts me to tell. One time - this goes back, oh it has to go back like 15 years -I remember at a conference, sitting around a table after dinner with, it had to be five or six people that were all really leading XML gurus, right? We were all just really involved with XML. And as we were talking about our backgrounds, it turned out that all but one of us were English majors. Every one of us but one. And the guy that wasn't an English major - one of the most brilliant people in the industry, his name is Evan Owens - was a musicologist. So people are thinking, "Oh, these must be all computer science guys." It's like, "Oh no, not really. We're all English majors."
Len: That's great. What was your first job in publishing?
Bill: I worked as a proofreader, actually for a music publisher. I am not expert in music, so this is an interesting little anecdote for you. I was a really excellent music proofreader, because I couldn't read the music. So I didn't know what the note was supposed to be. The danger in a proofreader is they read what they think is supposed to be there rather than seeing what actually is there. I have a real eye for detail, and so I could notice things like, "Wait a minute. That's not the same as that thing, and it should be." I had no idea what those things were.
I was also a graphic designer by background. My boss at the time discovered that I was a talented designer, so I got designing things. And then ultimately we bought typesetting equipment really early on, when it was all going electronic.
I ended up running my own company. For many years, I ran a publishing, design, editing, and typesetting publishing services firm.
Len: Was that Impressions Book and Journal Services?
Bill: Yes, how'd you know about that?
Bill: Oh, duh. I forgot that that was in there. Yes, absolutely.
Len: Part of that name is "Journal Services," and I know that you've got quite an extensive background in scholarly publishing.
Len: These kinds of high-level questions are kind of hard to get right, but how have things changed in the last, say 20 years in the scholarly publishing space?
Bill: Complete transformation, basically. Scholarly journals were one of the first sectors of publishing to move away from print and to move online.
Now, admittedly they were doing that with PDFs, not with HTML websites at first. What used to be very traditional publishing workflows and practices that were designed around print workflows - a lot of paper, editors marking up papers with red pencils, typesetters retyping the content that they got - the author sends a manuscript in as a stack of paper. And then it gets edited, so it gets all marked up. And then that gets shipped off to the typesetter , and they sit down at a keyboard on some other kind of machine, and type it all in again. And then somebody has to proofread it.
All of that work became electronic. Word processors came along, like, "Hey, we could use these Word files." Guess what? We can actually edit with the Word file. Now we can transform that Word file. Rhe Word file now actually is XML under the hood. So if you've got a Word file, you've actually got an XML file right there. And that can feed into a production technology that produces a PDF - but it also can produce HTML for online, or it can produce an EPUB.
EPUB is not used much in scholarly yet - except, potentially for scholarly monographs. A lot of university presses do EPUBs. But journals, which are the majority of scholarly work - scholarly literature is journal work. Those are single documents, typically. So they're either online, a web page, or they're a PDF.
My focus is publishing technology. So that's really the perspective that I'm coming from. If you're asking me how journal publishing has changed, the business side is dramatically different too because there's this rise of - originally, scholarly societies would publish journals in their specialty. Gradually more and more of that got taken over by big commercial publishers. Then, there was pushback on that.
Now there's a big trend toward open access, which means that the cost of production of the published work is actually done up front by what are called Article Processing Charges - APC's. And then the resulting content is freely available on the internet, behind a subscription paywall. So that's a big change too.
Len: There's so much to talk about in this area. I saw a talk you gave on YouTube - well, you didn't give it on YouTube, I found it on YouTube.
Bill: I'm not even aware of those things being out there. Every now and then somebody says that and I think, "Oh, really?"
Len: There are a few, often from conferences where you've spoken; people have recorded them and uploaded them. In one of these talks, you describe how the inherent requirements or characteristics of scholarly publishing led it to have quite advanced standardization, compared to things like magazines and newspapers, particularly the requirement in scholarly publishing for things like citation. You need to know what's being referenced. That needs to be a thing that needs to be identifiable.
I was very curious - I'd thought about that a little bit before, having a little bit of a scholarly background myself - but I hadn't thought about how the Associated Press puts out 250,000 new sort of things a day.
Bill: Oh, you're one of the few people that know that statistic. I quote that all the time, it blows people away.
Len: I confess, I didn't know it independently of watching your video. But it's really amazing. I was wondering if you could talk a little bit about how that works in scholarly publishing. What are the details of how things are defined?
Bill: That's a really good question. This is really why interoperability is such an important issue, and why standards are so important, as you started out saying.
Let me give you an example. The citation is a real driver in scholarly publishing, obviously, for a couple of reasons.
First of all, a typical scholarly journal article will have dozens, or maybe even hundreds of references at the end. Each one of those points to an article published somewhere by some publisher. Back in the day, those were all behind paywalls. So unless you were at a university, and your library subscribed to that journal that that article was in, it was very difficult for you to actually get to that article.
Another thing that characterizes scholarly publishing - and again, this kind of thing happened by necessity, but in the end, it's an altruistic thing - is that instead of some commercial entity saying, "We're going build a solution here and control this process," the publishers realized that no one publisher could, in fact, own and control the process of doing this interlinking between published content all over the world.
So a group of the leading scholarly publishers, which were and are fierce competitors with each other, banded together, and created an organization called Crossref, and used a technology that was just then being developed, called the DOI, the Digital Object Identifier, to build an infrastructure that provided this kind of interlinking between citations and the articles being cited. It's completely indispensable to scholarly publishing. If your article doesn't have a DOI and isn't in the Crossref system, it's invisible.
Basically the DOI is a dumb number, it's just an identifier. It's a unique identifier that is associated with a particular article. That unique identifier is what enables a system to - in the Crossref system, there's a ton of metadata associated with that identifier. Who's the publisher, who are the authors, who's the corresponding author, when was it published, what issue, what page? All that, the system knows, so that it's able to disambiguate things.
Now we've got identifiers for people too. The ORCID is a contributor identifier. So there are a heck of a lot of "Mary Smith's" in the scholarly publishing world, right? So just saying about this article, "Mary Smith is a contributor to this," is a huge problem for publishers and libraries, because they need to know, which Mary Smith? I have an ORCID. It's in my email signature. I'm not a scholar, I just like promoting this. So I've got an ORCID and it's in my sig.
Now there's a new standard called "The Credit," which will even say, "What was this contributor's contribution to this article?" Because sometimes there's a dozen different authors listed. Well, one or two of them might have been the main researchers. One of them might have been a translator or something. Or one of them might have been a tester. Now you can actually say, "What did they do?" All this is based on the fact that there's a community that comes together to agree on a standard and say, "Okay, we'll use this system and we'll all use it, nobody owns it." That's really cool.
Len: That is really cool. It's so interesting too how the problems multiply, because in addition now to things like - I mean, so far we've kind of mostly talked about attribution of author and publisher and contributor. But then there's the question of rights.
Len: You talk in one of your talks about how a textbook chapter, if it's also got an interactive kind of element to it as well, because it can exist in multiple formats at once - a single chapter of a textbook might have multiple contributors, including students. It might have a picture or pictures. It might have video. It might have clips from an interview from the BBC. It can have any number of things. How has the industry tackled the attempt to sort of standardize or automate, I suppose, rights allocation?
Bill: The answer is, how is the industry dealing with this? Because it's not past tense, it's a work in process.
Really the pioneering work on this was done in the news industry for that reason that you mentioned. The AP, for example - creating quarter of a million assets a day, 150,000 of those are text assets and 100,000 are image assets. You cannot process the rights manually for that kind of a fire hose of content. So they developed a specification called RightsML.
It became part of a group within the W3C - a community group that developed a markup scheme called the Open Digital Rights Language, ODRL. That was kind of an informal standard that was used by the news industry organization called the IPTC, which I'm also a member of - the International Press Telecommunications Council. They were instrumental in helping contribute to this work on rights. And now it's actually an official formal W3C recommendation - the ODRL, the Open Digital Rights Language.
This provides a standard way to express things like a rights holder, obligations associated with a piece of content. So, "You can use this image, but it has to have this credit line adjacent to it." Or, "You can put this article on your website, but not before September 15th, 2018" - an embargo or something like that. All the kinds of things that you have to manage. Or, "You have to pay me 100 bucks." For all the things you need to express about rights, ODRL provides a framework.
And then the news industry has developed on top of that framework, RightsML, which addresses what the news industry needs to express about I've been trying to socialize this - in other words, make other industries aware of this development. Because the trade book industry doesn't need to express rights in the same way the news industry does. Media is actually ahead of the book publishing industry in doing this. I have a colleague in Publishing Technology Partners, Bill Rosenblatt, who's been one of the leaders in this whole rights area, one of the leaders in the rights area.
Different sectors and different players can make progress on something because it's a necessity to them. And it turns out to be useful to everybody else. That's really a cool dynamic.
Len: This is a bit of a sideways question. But speaking of your colleagues, I interviewed Thad McIlroy for this podcast quite some time ago now.
Bill: Oh, that's great.
Len: And he had a wonderful story about being one of the first people to fully create a published book using desktop publishing tools.
Bill: He was way ahead of the curve on that.
Len: He has this great story about how Apple called him, because they found out, and they're like, "How did you pull this off?"
Anyway - my brother and I have this joke that there's an infinite number of ways you can divide the world up into two types of people, and one of those divisions is, people who were sort of carrying on in their careers, and then a computer was dropped on their desk. And they were like, "Oh my God, it's the end of civilization." And there are other people who are like - they saw themselves in the origins of pre-history at last,
What was your first experience with that type of technology, and how did you respond to it - if you can recall?
Bill: Oh, I definitely can recall. I mentioned that early in my career I was with a small publishing company that - because I was interested in getting into the technology, we actually got typesetting. So we actually had - get this, you will laugh at this = we had a word processing machine. It was called a Redactron. It was about as big as a washing machine. And it would record your keystrokes on magnetic tape, cassette tapes.
Then there was a typesetting technology from a company called Murgon Fowler Linotype [?]. And others - Compugraphic was another one, where it would actually typeset from those digital files. Originally it was phototypesetting - it was actual film exposed by light, and then it became digital. So anyway, long story short, is that this became very specialized, very expensive.
Fast-forward to when I had my own company - I would get a new version of the typesetting system that we used, and I'd get a couple of workstations, and they would cost me $40,000 a pop. Very specialized tools. But did just beautiful typography, very sophisticated.
So along comes desktop publishing. I remember there was a consultant in the field back then, Jack something? Thad would remember his name. I remember him saying, "Oh, this desktop publishing, it's a flash in the pan. This is just crap. It produces hostage-note typography. It's not a real professional tool." Well, guess what? It caught up. And then pretty soon everybody was using InDesign and Quark. And those big expensive typesetting systems that we'd invested a million bucks in, were kinda out of date.
Know something? They still are used. For example, a lot of journals are produced by those high-end systems, because they're just incredibly fast and incredibly automated.
Len: So it sounds like you were excited to embrace the new technology as soon as it became available to you?
Bill: I was excited, but I was also scared as I could be. Because I was an entrepreneur. I was a small business person. And it's like, "Man, if people can just do this themselves in their dining room, is my business going to fall apart?" But of course that is why I've always kept up with technology. You can't just sit there doing what you're doing. You'll be left in the dust.
Len: That's a great story, thank you for that. You're reminding me - when you bring up the consultant who's like, "This is a fad," which of course people said about TV, and there are, to this day, people who say that about ebooks.
I know it would have come in kind of stages, through various types of devices, but when you first encountered an ebook on an ebook device, what did you think about that?
Bill: Well first of all, I thought it was the future. Back when I had my company, I basically gave two of my employees a lot of support in getting involved with the development of the early standards. It was called OEB, The Open Ebook standard - which was the predecessor to EPUB. I've just always been devoted to standardization of things, right? Because it just makes everything work better.
So this was pre-Kindle. 2007, when the Kindle came along, was a watershed. Because now suddenly, it was really a mass-market thing. Before that, it was like kind of a Tower of Babel. There were lots of different ebook reading systems, but they were all proprietary and they all worked differently. So, Kindle is proprietary. But at that time, when the EPUB standard came along, pretty much all reading systems except Kindle came to basically be able to work with EPUBs natively.
Even to this day, even though Kindle is stubbornly not in EPUB, the format that Amazon wants you to submit your content in is EPUB. What that means is that a publisher can create one file that they can send out to all these different reading systems, aggregators, retailers, etc. - and that one file can be used by all of them. It's not used the same way by all of them. They all have their bells and whistles. They'll do things differently.
Apple, for example was one of the first ones that if you properly tagged a footnote as a footnote, they would do it as a pop up. That was very cool. Because you used to have to lose your place in the content to get to a footnote, and then you had a hard time finding your way back to where you were. And they just made it pop up. The reason I bring that up is that this kind of standardization does not squelch innovation, it enables innovation.
Len: Researching for this interview, I thought a lot about standardization in ways I hadn't in the past. It's really fascinating. I'm sort of jumping ahead a little bit here, but there's not just a difference in degree, but a difference in kind, when there's more than one standard for things - when there are multiple ways that you have to kind of do things.
Len: Do you think that Amazon will always keep its MOBI format, or do you think that they might someday acquiesce?
Bill: I was going to say, "Confidence." I should probably say, "Optimism." But yes, I think they are moving in that direction. The fact that they now prefer to ingest EPUB 3 as the best format to provide them content is a good sign. They actually have a big commitment to accessibility, which really depends on EPUB 3. EPUBs are just becoming more and more pervasive.
For example, the latest version of the Edge browser will just open in EPUB. You don't have to have a separate EPUB reader. It'll just open it up like you can open up a PDF. That's pretty important development.
Len: I'm really looking forward to asking you some questions about EPUB in a moment. But before we do that, you had a recent article published in Publishers Weekly, in which you say you have a story about meeting Jeff Bezos at, I think the Library of Congress, back in the day when he was pitching this wild idea to use the internet to sell books.
Bill: That's right.
Len: I was wondering if you could tell us a little bit about that story?
Bill: It was a conference that - what was her name? Jane something. Sorry, I should remember. She was at the Library of Congress. She was actually at a university in the DC area, and worked with the Library of Congress to do this publishing conference every year in the spring. I spoke at it a number of years. And one of those years, she brought in this guy named Jeff Bezos to be one of the speakers. He was a really dynamite guy. He was an outsider - he was not from publishing.
This is another interesting observation, that oftentimes the best innovations come from people that are not already embedded in the system. Because sometimes you can't see outside of your context when you're so embedded. I think he was an investment banker by background.
Bill: And an entrepreneur. He just saw this technology and just saw the potential of it. And of course, he has a driven kind of personality. So part of it is, you need to be smart enough to see these things. But you need to have the personality that can actually make it happen. Because lots of people have lots of good ideas, that never actually go anywhere.
Anyway, he had the audacity, and he made a lot of enemies in publishing. Before they loved him, they hated him, I think is a good way to put it. Because he upended the whole book retailing ecosystem.
Len: I think he might still be hated in some quarters.
Bill: Oh for sure.
Len: You may have seen the news recently that a book that an author published through Amazon's CreateSpace service was put on the nominated list of books for a literary prize in France. A group of people representing book stores and the book industry in France got together to protest against this. There's some technicalities involved in this, but essentially they would have to buy books from CreateSpace to put them in their store to sell them. And they characterized this as forcing bookstore owners to contribute to their own demise by feeding Amazon's bottom line. Just at a high level, what's your take? Is Amazon a force for good in the book world?
Bill: Well first of all, thank you for that anecdote, I had not heard that story. I'm not sure what kind of reach your audience has. So, if people are not familiar with France, the bookselling sector in France is very, very different from the US. The government actively protects bookstores, actively mandates that all the books have to be sold at the same price. It's a very, very different environment than it is here.
But - and of course, we're in publishing, we're all about books - but they were just a stepping stone for Jeff Bezos. He sells diapers now. He sells anything. That was part of his brilliance. He just thought, "Well, wait a minute, books are easily available, easily packed and shipped, easily identified, there's a ton of them. I'll start out just selling those." But he always intended to sell everything.
I have a strong resistance to binaries. So to say, "Is he a force for good? Yes or no?" - it's something that's very hard for me to respond to. But to say, "What good has he done?" I can say - he has really made a very tradition-bound and stodgy industry suddenly sit up and rethink how they do what they're doing, and modernize workflows and bookselling. It is still true that for trade books, the trade book industry is still very dependent on Amazon.
For Amazon, I shouldn't say "he." Amazon has a reputation for playing hardball and really being an unfortunately dominant force that has way too much power in the industry. But on the other hand, some of the innovations and developments, and making publishers respond to this new environment. are very good things.
You mentioned my colleague Thad. One of the things Thad is really a specialist in, is metadata and optimizing metadata. If you're a trade book publisher, optimizing your metadata so that you can sell more books, is always an important goal. And most publishers don't do that very well. But the reality is that they mainly want to sell them on Amazon, because that's mainly where people are buying them these days.
And we're not talking about ebooks necessarily, we're talking about print books. It's just - talking about diapers.
Len: Yeah, it's fascinating. That's where all the eyeballs are. Even if you're in other places, you need to be there too.
Bill: It means that people that aren't in that space still are influenced by that whole development. So, like the work that you do at Leanpub. You've got a very advanced platform and distribution model, etc. But I mean, obviously, virtually. I don't know a lot about your background. But looking at your website, I would say that - had Amazon not come along, I wonder if Leanpub would exist? They altered the environment, so something innovative like what you're doing can have a real compelling place in it.
Len: For sure - Leanpub was founded in 2010, and the Kindle had already been around for three whole years by then.
It's interesting, I haven't been around all that long. But even just in the last eight years, the normalization of ebooks in particular, and even the normalization of people's understanding of the internet and computers, has changed noticeably.
We just don't get the same kinds of questions that we used to get from people. People know how this stuff works, and they know what it is.
Bill: That's really interesting.
Len: It's maybe once every six months that we get someone complaining about wanting a "real" book. That used to be more common. People see ebooks as real books now.
And they know how they work. They know how to get things on their Kindle. That's something that we used to have to do a lot of work to convey to people. Now it's not so much that everybody knows, but when we when we point them to instructions, they have the vocabulary already to understand those instructions.
And 100%, without Amazon having introduced the Kindle and having really put its weight behind ebooks, the world we live in would be very different right now.
Bill: That’s right. Another thing that that makes me think of, is that - I told you that I'm very distrustful of binaries. So, there are different kinds of books, there are different kinds of markets. Trade, that's kind of plateaued at about, what? 10%, 15% maybe at best, ebooks, and it's been there for quite a while. But I noticed on your website you publish a lot of technical books. And those people don't want print. Academic books, libraries - we talked about journals going electronic long ago, but now in terms of scholarly books - that's moving in that direction too.
University presses love beautifully-typeset and beautifully-printed books that they sell 200 or 300 copies of. But libraries increasingly don't want the physical book. It also depends on the nature of the book, because there are a lot of books that aren't meant to be read, they're meant to be consulted. That's a whole different thing.
Len: There's a lot I could say about that.
Bill: I bet.
Len: One thing, actually - you reminded me when you brought up libraries just now. I think that it's easy for those of us who aren't in the standards world to not understand how difficult it is to set and implement standards, and how important they are.
One example of that - the fact that, knock on wood, the building I'm standing in right now isn't falling down around me - is because of standards.
Bill: One of my customers by the way - I did a lot of work for the International Code Council. That's who makes those building code standards
Len: It's so interesting to know that the breadth of this this -
Bill: Just walking up a set of stairs - it doesn't occur to you that the stairs won't be evenly spaced, and that there'll be about this much to go from one step to the next. That's a standard
Len: And realizing that like engineering, standardization is its own profession. Which I mean, to you, would not be news.
But to a lot of other people, the idea that it's actually a category of activity, and that it can apply to basically anything, and that the reason our world works as well and as poorly as it does, is because of people getting together with this set of ideas that's evolved over time. I presumably really got going in our contemporary world with the Enlightenment and things like standardization of - but standardization of weights and measures of course is an ancient thing, and exists in all kinds of cultures that developed it as an idea, independently of each other.
But maybe zeroing in specifically on libraries. So I remember back in the day when LaserDiscs were a thing. I only know this from the outside, but from what I heard around the universities, was that a lot people were investing a lot of money and time in getting all their data into LaserDiscs. And of course that didn't last. Particularly with libraries, I'm going to sort of set them out like they're - particularly in the scholarly world, they play a profound role in our societies, of preserving things. How has that requirement affected discussions in the past about e-publication standardization?
Bill: Huge issue. You mentioned that I did some work with the British Library. It was actually exactly on that subject. Because at the time - the British Library has always been basically the depository of published content by British authors. Just like the Library of Congress for us. All published books that get copyrighted are supposed to be sent to the Library of Congress, and they keep them. They preserve them.
Well, preserving print books is one thing. What happens if there is no physical object? What I worked with the BL on was analyzing the kind of e-journal landscape. Because it was kind of at an inflection point, where it wasn't just that journal publishers were grudgingly offering a digital file in addition to the print, and were suddenly saying, "We don't need just print anymore, we just want to send the digital file." And the library is saying, "Good, because we don't have room for all this print. We just want the digital file."
But now, how do we preserve this? Because - I was going to make a comment, and I want to come back to this - but one misconception people have about standards is that they're stable. They actually change constantly. So no we don't use LaserDiscs anymore, but the file formats change. So even now in the W3C, we're about to bring out a new version of EPUB, called EPUB 3.2. And the main driver of that is that when we came out with 3.1, there was some big improvements and simplifications that we made about the standard beyond the EPUB 3.0.1 - that was the previous version. But it was not backwards-compatible with that previous version.
Well, that was a problem. Because now there's all these older EPUB 3s out there, and any reading system now has to deal with two different kinds of specs at the same time. So EPUB 3.2 captures the innovations in 3.1, but presents them in a way that is in fact backwards-compatible with 3.0.1.
I mentioned being involved in these standards organizations, in things like ASME, which is engineering standards. Or ICC, which is the building code standards. They are constantly in flux. The boiler code for the ASME has something like 3,500 engineers around the world constantly working on that standard. Most of these standards go through, typically, a three-year revision cycle. It's because things come up, and they have to address things that on the one hand address some new issue that's come up, but don't break what's come before.
So anyway - I got myself off on a tangent there, but I thought it was an interesting tangent.
Len: It is. One thing I wanted to ask about was - if you could describe how EPUB 3 is managed, and maybe just explain what it is, because there's an entire organization behind it.
Bill: That's right. First of all, the EPUB standard, for many years, was governed by the IDPF - which is the International Digital Publishing Forum, which, basically was the successor to the open ebook forum that made OEB - it became IDPF. It was a membership organization that publishing companies, technology companies, etc. would join, would band together, and send people to these working groups, to work out the standards. Most of them are very, very consensus driven.
So it's not a matter of, there's two possible ways of accomplishing X, which way do we decide we're going to do? It's not just a matter that there's 30 people in the room, and 16 of them vote for one and 14 of whom vote for the other, and so the 16 win. That's not how it works. They have to talk about the merits of way A versus way B, and ultimately come to consensus on it before it gets put into the standard.
So anyway what was happening - the ebook standard, basically the EPUB standard - creates a single file. Many people think it's just a file, because it's a .EPUB file. But it's really a package that contains a whole bunch of content documents, typically chapters in a book, etc. It's got fonts, it's got images. It may even have videos, or it may connect to a streaming video. It's got a lot of metadata, etc. That whole package is what is actually an ebook.
That standard over the years has gotten more and more completely aligned with open web technologies, the open web standards of the W3C. And at the same time, the W3C realize that as sophisticated as the web technologies have gotten, they really don't handle what is a true publication. A web page is not the same as a complex publication. It's a thing, and it can link to lots of other things. But those other things it's linking to aren't part of it. They're external to it.
Whereas, a publication packages everything up, and everything is has an identity. So they were realizing that they needed to begin work in the W3C. to address this issue. And so here we had the situation where two separate bodies were potentially working on the same thing, and might come up with two competing standards, which would have been a disaster for the industry.
Len: If I can just jump in for a moment.
Bill: You can.
Len: We've talked about the W3C a bit, and I just wanted to mention for anyone listening, this is like the Worldwide Web Consortium, a very important organization.
Bill: Yes, they govern XML, HTML, CSS - all of the standards, hundreds of standards that are involved in in the web world today.
About a couple years ago, give or take a few months, the IDPF became part of the W3C, so all that work is now happening within the W3C.
There are three components of the organization, if you will - under the umbrella publishing at W3C.
One of them is a publishing business group, and that's an advisory group - non-technical people that are responsible for communicating the needs of the publishing industry to the W3C. Again, we're not just talking about books, but scholarly publishing, educational publishing, technical publishing - all those kinds of things.
And then there is the Web publishing working group, which is working on this vision of a web publication. The vision there - it's actually more than a vision, because the work has gone ahead quite robustly on this - so that right now, you can either have a web page, or you can have an EPUB, but they're not the same thing.
What we want is to be able to have a publication that either just lives on the web and is online, or is packaged and can be used offline, or is cached, for example. So technically it's offline, but it's not packaged. Getting into the technical weeds there.
But the idea is, it's just one thing, and why that's so important, is that for a publisher, this would be nirvana - that you could produce this one file that somebody can just open up in a browser. They can open a reading system, they can store it on the hard drive. They could store it in iCloud if they want, or whatever. But it just works in any of these contexts.
That group is called a working group, and they publish the official standards that are called recommendations in the W3C. So this web publishing recommendation is due to be published, at least its first incarnation, in about another year and change. But along with that, they're also chartered with spec'ing a packaged web publication. And they're also chartered with spec'ing EPUB 4, which is the next generation of EPUB, which will be a type of web publication.
The example I usually give to people - and again, keep in mind I'm speculating. I'm involved in these working groups, but the work is ongoing, so the decisions haven't been made - but typically, web publishing wants to use any technology that's a valid web technology. So there are a number of different packaging methods that are valid web ways of packaging stuff. Probably a packaged web publication will use any of any of those, whichever one you want to use.
For something like the book supply chain, or publishing supply chain - we're back into the messy unpredictability that we had back in SGML that I was talking about - the supply chain wants to know that it's packaged this way. Right now an EPUB is packaged as a zip file.
An EPUB is really just a zip file, a zip package. What that means is, if you get one of those things and you need to unpack it, you know exactly how to do that.
The third group within W3C, is the EPUB 3 community group. That's a group that is charged with maintaining EPUB 3. Yes, EPUB 4 will be aligned with this web publication, etc. But we've got now a well-established, big ecosystem using EPUB 3, so we don't want to pull the rug out from under that.
Len: Sorry to stop you there, but I think it's a good moment to reflect upon how historically contingent the success of EPUB 3 actually is. In one of your talks, you mentioned that there was a problem when EPUB 3 was first announced, which was that people would complain, "Oh but there are no tools to use it with yet." And you're like, "Well, that's because it's a new standard and you want to have the standard before the tools." But it creates this inherent kind of paradox, to actually having a standard be adopted.
Just from someone who would know - what was it like when you heard that Apple was going to use EPUB for iBooks?
Bill: Apple was on the working group.
Len: Oh, they were.
Bill: To create some standard. Yes.
Bill: Even though they're as secretive as Amazon, they play all their cards very close to the chest, it's very hard to get an Apple employee to talk publicly about anything they're doing - back when EPUB was being created in the first place, they were a very major contributor to it.
One example that I always like to cite is that we were working on a spec for synchronizing text and spoken audio. That's really important for accessibility. In fact, the technology that does that comes out of the accessibility world. So that was being worked on as part of the discussion of the EPUB spec.
And then Apple brought out iBooks with Read Aloud. Your first thought would be, "Oh dammit, now there's this commercial standard that's proprietary." Well, it turns out that they basically used exactly the spec that became media overlays in EPUB. So there were a good corporate citizen in that they did it the right way. They knew what the spec would be, it's just that they came out ahead of the spec in that case.
That brings up another really interesting dynamic, when IDPF became part of the W3C. Because this problem that you cite is a real problem, and another aspect of this problem is that there have been things over the years in the EPUB spec that have never been implemented. There been things that everybody thought, "This is a really great idea. We'll put this in the spec," and then nothing ever happens with it. And now what are we going to do?
The W3C takes the opposite approach. A spec is not final unless you can demonstrate, I think, two independent implementations of every feature in your spec. And at the time of balloting - when it gets to that late stage of voting on whether to make it a spec or not, I'm BISG's representative to the W3C - see so I actually get to vote on these things. Fefore it gets to that last balloting stage, there is this stage where implementations are being invited and sought. And if it turns out that a particular feature of the spec has not been implemented - and it has to be two independent implementations, it can't be two variants of the same implementation - it can't be in the spec. The spec gets revised so that gets pulled out of the spec.
Once it reaches its final recommendation stage, everything in there has been implemented somewhere that does two things. It demonstrates that it can be implemented, so that we haven't spec'd something that's just impossible to achieve.
But the other, more subtle thing is that somebody thought it was worth the trouble to implement the darn thing. So it demonstrates viability, and it demonstrates interest.
I think those of us involved view this as a very positive consequence of now being part of the W3C. There's much less danger of things getting in the spec, and now the technology out there doesn't just catch up with it.
Now, does it mean all the technology implements it all right away? Of course not. Some people will never implement x, y, z features. So even something as basic as well, the read aloud, the media overlays. Not every system can do that, but there are a lot of systems that can.
Len: Thanks very much for that great story. It's really really interesting to hear how the sausage is made. Sorry for the cliché, but, but it is really interesting. It's also a good reminder to understand that behind every technology, is a lot of people getting in fights, taking initiative, being concerned.
On that note - I'm getting the impression we could probably talk for a few hours about these kinds of things.
Bill: We could, yes.
Len: But we should probably move on, because I definitely want to talk to you about accessibility.
But before going onto there. just as a sort of bridge - today happens to be the International Day Against DRM or Digital Rights Management, for those listening who might not know. I don't know if you've heard of this this event. it happens every year. It's put on by a pretty small organization, but we like them.
Just generally speaking, for those for those listening - Digital Rights Management can mean a lot of different things. One form it can take, is someone can't just send you a file, and you can open it. You've got to have some kind of key - like a password or something like that, to get in - to put it in a sort of simplest form.
This is, I think, still a controversial issue in the publishing world. Let's say in the ebook world, let's narrow it down to that. What's your view, Bill, on DRM?
Bill: Well, let me just say that because I operate in lots of different spheres - and because, frankly I'm a consultant, so I need to be very flexible, to adapt to the needs of a particular situation.
So, I am neither for nor against DRM. Commercial trade publishers, absolutely, are devoted to DRM, because they need to make sure that their content isn't pirated. That's the problem with these digital files, is they're infinitely reproducible in perfect form. So once you've got one, you can make any number of copies and send them out. Nobody's paying for them, and now you've got a problem.
In fact, one of the things that Bill Rosenblatt is looking at, I wouldn't say "working on," but he's actually chairing a session at Digital Book World about this in a few weeks. There's a couple of blockchain initiatives that are addressing this issue, so there's potential future there.
But to take the opposite approach, from the accessibility point of view, they absolutely hate DRM. Because those EPUB files that they get, that otherwise should be really accessible to them, they can't even crack them open, because they're protected by DRM. So they have to get a DRM-free file.
So there's whole systems in the accessibility community, Benetech is a leader in this - with a service called Bookshare, where they will get the content, get the books from the publishers. They've got half a million of them at this point. They will provide an accessible DRM-free version to people that require them - blind people, people with learning disabilities, cognitive and visual disabilities. etc. But "print-disabled" was George Kerscher,'s term, and the prevailing term for that. The key thing that they do is they absolutely ensure that it's only distributed to people that are qualified to receive them, and they can't be distributed outside of them. So that's how that gap is gotten around.
From a standards point of view, EPUB has nothing to do with DRM. That was deliberate from the very beginning. It neither enables nor prevents DRM. Web publishing, I guarantee you, will be the same thing. In the Web world, I refer to that as the acronym that cannot speak its name. Because in a conversation, in a meeting of techies, when the subject of DRM comes up, the conversation grinds to a halt. Because everybody in the tech side hates DRM. Commercial publishers depend on it. So the standard itself has nothing to do with the DRM.
One misconception that a lot of people have, is that the publishers are putting the DRM on the files. No - it's the distribution platform that puts the DRM on. A publisher's EPUB doesn't inherently have any DRM on it. But when they send it to Apple for iBooks, Apple puts its DRM on. And when they send it to Amazon for Kindle, Amazon puts its DRM on. So here's the same book. It's got two different DRMs. Why is that? Because Amazon wants you to only use the one you buy from them in their system, their ecosystem. Same with Apple. Same with Kobo, etc. So it's really not the publishers that are putting the DRM on. It's the people that are selling you the book, who are putting the DRM on. It's not inherent in the system at all.
Len: Thanks for that great answer, and the way you so well-articulated the very different interests behind people's embrace or rejection of something like DRM, and that it's a real issue. You might get the suits in the room, and the people who know how the computers go in the room, and they might have different interests, but it's important - whether you object to the other side or not, it's important to understand deeply where they're coming from.
One thing on the subject of accessibility that I was just reminded of, was - I saw a while ago, actually an old now documentary on the history of the computer, and there's this very moving scene, when the documentary makers film a group of children, each of whom has their own disability or disabilities. And they're using these computers in various ways, in order to read for the first time on their own.
There was something about being able to turn the page, as it were. This is on a computer. But there was still - just being in control of the text it itself, was a revelation, and a wonderful thing.
Accessibility can mean a lot of different things. I mean, I grew up in the days before the internet, in the prairies of Canada, in a sort of lower-middle-class family. And I know what it was like to not have as many books to read as I wanted. Now, thanks to computers and the internet, I've got a lifetime of books to read just on the Gutenberg Project alone.
I wanted to ask you, how did you become passionate about accessibility? Was there something in your life that prompted it?
Bill: First of all, I saw how important it was. As I'm participating in these standards groups, it's - the people from the accessibility community are very active participants. Because obviously it's in their interest for these standards to be made in a way that they're accessible, and that is in fact what has happened.
So no, I don't have any kind of personal - well, I do have lots of personal acquaintances now. I have so many good friends in the accessibility community. I mentioned George Kerscher earlier on, and he's he's just a fabulous guy. He's blind, he's got this wonderful dog named Croner [?]. He comes to the conferences with Croner, and Croner just crawls under the bench or under the seat. And George is this amazingly brilliant and articulate guy and a sweet guy.
So anyway, he just does everything to help make the world more accessible. I admire that greatly. So yeah, I know a lot of people that work in the area. I've become passionate about the issue, but I can't actually say that there was some kind of triggering event in my life that did it. It was just a matter of a growing realization about, "Wow, this is really important. We should be paying attention to this."
Len: I remember in one of your talks, you describe in quite sort of strong terms - this might be better now, but that terrible situation that university accessibility departments found themselves in, where they would have to like tear the spines off books in order to, I think, take pictures of the pages, in order to distribute them to students who needed that format for accessibility.
Bill: That happens all the time to this day. It's still shockingly common, and that's something that we're doing a lot of work on.
I mentioned my committee in the Book Industry Study Group. One of the working groups is specifically trying to address that issue of working with - these are called DSO's, which are Disability Services Officers, or DSS offices - Disability, Disabled Student Services Officers. Most universities and colleges have a unit like this. Oftentimes it's one or two staff members, and some interns.
I really admire these people, because, first of all, the university is legally obligated to provide an accessible version of course materials to a student who needs them. And so these people and these DSOs, or DSS officers, have to scramble to get some kind of file to work with. And if they don't get a file that they can work with, they will literally go to the college book store, and buy a -
Jamie Axelrod is a friend of mine. He's the immediate past president of AHEAD, which is a national organization of these offices around the country. And he's head of that operation at Northern Arizona University. In his talks, he'll show a slide of a guillotine. You know what a guillotine is?
Bill: It's what is used to chop the spine off the books. It's sitting right there in their office. They don't use it every day, but it's a daily presence. They use it all the time, where they have to chop the spine off the book. They then feed the pages through a scanner. And then they do OCR, which is Optical Character Recognition, which is imperfect, and which doesn't capture any structure or tagging. And then they have to go through and proof read it and tag the book, etc.
It drives me crazy, because these were all well-structured digital files that they typeset the damn book from in the first place. way upstream, and they just can't get those files. Or, if they had just gotten an EPUB of that book, they'd basically be 90% of the way there.
Anyway, they're doing heroic work, but it's really still a very messy situation. And oftentimes, that book that gets remediated at Northern Arizona University - what are the odds that there are half a dozen or a dozen or more other colleges and universities doing the same thing with the same book - because they've got a student that needs that book.
I'm working with a group that's - it's too formative to talk about right now, but to try to address this situation, so at least if one university remediates a book that another university needs, there's a way that they can know about that.
But right now, it's just shockingly redundant work, and shockingly manual work, that just should not be happening. So yes, I am passionate about that.
Len: I'm sure a lot of people are very glad that there are people like you, and people in the groups that you work with, that are out there trying to improve this.
Bill: Yes, lots of people working on this, absolutely. I'm just one of them - one of many.
Len: To end on a positive note, you published an article recently in Publishers Weekly, that I mentioned earlier, called, "A Golden Age of Publishing Technology."
Len: Could you talk for a couple of minutes about what makes this a golden age?
Bill: Well first of all, what I said was, "We are approaching a golden age." I didn't say we were there yet.
Bill: I did that fully realising that people will interpret it that I'm saying that we are in the golden age. And actually the editor changed my title a little bit to imply that. It was "approaching a golden age."
Len: You don't get to pick your own headlines.
Bill: Not necessarily. Oh, but PW is just great to work with.
Bill: Jay Milliad [?] at PW is a terrific guy. They changed almost nothing in my article.
What I meant by that is that - this goes back to the beginning of our conversation, where I'd lived through that whole history of handset type and rub-on letters, phototypesetting, all that kind of stuff - all that stuff was laborious, and it got more and more automated - but even so, was using proprietary tools. They didn't talk to each other. If you had a linotype phototypesetter and ot got out of date, and then you bought one from Autologic, you couldn't use any of the fonts you had before. You had to buy Autologic fonts for the new typesetter, not linotype fonts etc.
I was going to say, it's crazy. It's only crazy in hindsight, because it was just the way the world was back then. But what's happened over time is that gradually, because of standards, things are getting more and more interoperable.
I mentioned that the file that the author, or authors, write their book or journal article in, is a Word file. But it's actually XML under the hood. And so now that's processable downstream in the workflow, and nobody has to retype the stuff anymore. That's a trivial - well it's not a trivial, it's a significant example - but one of hundreds of things that are gradually more and more interoperable.
I mention now that one of my friends is Liisa McCloy-Kelley from Penguin Random House. She's head of ebooks at Penguin Random House. They produce a lot of very complicated ebooks, because they do children's books and cookbooks and stuff like that. They're to the point where they can produce one EPUB for the supply chain, and it goes out to all these different retailers. That's a watershed in publishing, to think that you can produce this one file - and you're using tools that are commonly available, and you're using standards that are well known. There's tons of people that can do HTML, for example, and CSS - a lot of people know how to do those things.
Another friend is Dave Cramer at Hachette Book Group. They actually typeset all of their books. This is the US arm of Hachette. They produce all of their books with HTML and CSS technology. There's no actual typesetting technology, other than - ot's called Prince XML that they're using. But it's basically the same technology that makes a website or an EPUB. It's not some different technology.
Luka Duran [?] in Hachette France, has a different approach. But they won an award from the Accessible Books Consortium this year for their work. What they've done is, they publish huge, huge numbers of books. So they are a big player, and they have a lot of clout in the marketplace. They've basically mandated that their suppliers have to supply an accessible EPUB as a deliverable for every book they publish. What that does is it puts pressure upstream in the supply chain. Now there's a whole bunch of vendors that can routinely do accessible EPUB's. Because Hachette made them do it, basically.
All of these dynamics are resulting in an ecosystem that has less and less friction, and fewer and fewer barriers to hand off a file from one system to another, one person to another, etc. And it's getting more and more interoperable, more and more widely available. A lot of the big developments now are happening in open source. So the software's even free in a lot of this stuff - a big, big difference. I would say that's getting pretty close to a golden age.
Len: Actually, to end on a low note, I've got another last question for you.
You've just reminded me about the importance of distribution streams and all the various pressures. What happens if Barnes & Noble goes under?
Bill: Well I noticed that you said "if."
Bill: You are expressing some confidence there.
Len: I was being politic.
Bill: You were being diplomatic, yeah.
To the trade book industry, to the trade book ecosystem, it will have an impact. But I don't think it will have any kind of a cataclysmic impact at all. They no longer are the main vehicle for publishers to sell their book: Amazon is. Kobo's doing great things internationally, for example. Indigo, Waterstones are doing some good stuff, etc. And actually the independent bookstore sector has blossomed in the past five or ten years.
We may end up thinking that the era of the giant retailers - Borders, Books-A-Million, Barnes & Noble, the three big "B's" - was a blip. We may get back to a day - uou can tell that I'm just an optimistic, positive person.
Bill: Sorry, I was just born that way. So here's another golden age little element that I hadn't thought about. A really good golden age would be - you can get every book you want in an instant digitally. But you've also got great bookstores in every town that are local bookstores, where authors come and do their readings. And those bookstores know their people.
I live in Ann Arbor, Michigan. And there's a bookstore called "Literati," that I like a lot. So here I am, a publishing technology evangelist, building the standards for digital books - and I still like reading print books. I just finished, Lincoln in the Bardo. But I bought it - I can't remember if I got it at Literati. Autographed book. It means that Saunders scribbled in the front of it. I loved reading that, the physical book. That particular book was a particular good one to read that way. I like the idea of, now, more specialized, idiosyncratic, personal indie bookstores, all over the place - complemented by this gigantic digital infrastructure. So everything's available, but it isn't homogeneous big box stores anymore.
Len: Well, thank you for taking my potential downer question, and turning it into a positive and optimistic note to end the interview on. And thank you, Bill, very much for taking the time to do this interview today, during what I assume is a beautiful afternoon in Ann Arbor.
Bill: It is actually a lovely afternoon. We don't always have lovely afternoons, but this one is a good one. That's right, it's been a real pleasure talking with you, lots of fun. And boy, I have to complement you - you do your homework. That was very impressive.
Len: Thanks very much. Thank you.