FeedbooksE-bookers should watch the IDPF carefully to make certain that neither Adobe nor any other company turns .epub into a proprietary standard through extensions or otherwise. But guess what one of the best countermeasures is? Use of .epub! And noticing suspicious situations and pointing out omissions, which ideally the IDPF can address in time—for example, lack of reliable interbook linking!

Readers and publishers of all kinds, then, not just public domain people but also Random House and little publishers, should thank the open-source-based Feedbooks site where co-founders Hadrien Gardeur and Loïc Roussel are fighting the good fight for genuine compatibility and setting a good example.

Tolstoy and scores of other writers already in .epub

From Tolstoy to Orson Wells and Creative Commons-licensed writer Cory Doctorow, Hadrien and Loïc now offer scores of authors via a beta treatment of the .epub format. Hadrien has invited e-bookers to sample his free wares, and, yes, when I used the easy-to-install OpenBerg plug-in for Firefox, things went smoothly. Page transitions via the space bar were a little sluggish, but that’s very possibly a Firefox thing rather than an OpenBerg problem. I may update this item when I’ve sampled the Feedbook offerings further. Bottom line: the .epub action now indisputably extends beyond dotReader and Adobe Digital Editions and other commercial products. (Update, 5:12 p.m.: I should have mentioned the wonderful FBReader, too, especially since it’s the program  I use on my Nokia 770.)

Congratulations to OpenBerg, of course—see OpenBerg-related threads at MobileRead. Yes, I know. OpenBerg in its present form isn’t the ultimate e-book reader; I don’t know if it plays well with CSS, for example. But it’s a great start, a wonderful way to begin to befriend .epub. I hope that someday Firefox and all other popular browsers can come with native .epub capabilities, or at least that many more .epub add-ons will show up for browsers.

Open source creation tool available as well

Hadrien also, by the way, has a linux-based open source creation tool for .epub. So, Project Gutenberg, shouldn’t you, too, keep an open mind about .epub—all the time looking for deficiencies? Based on its role in the jailing of a Russian hacker, not to mention a gigantic list of DRM-related sins, Adobe isn’t the most beloved company in the public domain and open source communities. But let’s care less about Adobe’s involvement and more about mechanisms for continually monitoring the purity of the standards.

Like it or not, big-time commercial publishers plan to use .epub to one extent or another. PG should involve itself, then, with what will increasingly be a mainstream format. Same for Matt McClintock at Manybooks.net. And if the IDPF is smart, it will cherish public domain people’s interest in the format. If the .epub survives the public domain community’s scrutiny, this will signal to publishers and the public alike that the format is trustworthy.

Needed: Credible tire-kicking

Few publishers and perhaps none really have in-house technical resources to completely vet .epub and related standards. This is one situation where crowd-sourcing ties will actually help commercial interests.

I also hope that demanding scholarly presses and scientific presses will fully embrace .epub and publicly point out any flaws, so that the publishing and library communities can encourage the IDPF to address them.

A “must read” from Jon Noring

Meanwhile I would urge TeleBlog readers to check out Jon Noring’s new post on the handling of footnote equivalents in .epub.

Jon writes that “we must not be constrained by following the limitations of print, but to transcend print and be able to digitally present content in ways better than print. Displaying annotations in popups, for example, is one such ‘better way.’ EPub certainly provides the framework to display out-of-spine annotative content in powerful ways — it is now up to reading systems to enable it and for publishers to take advantage of it.”

So will publishers feel comfortable with the handling or footnoting—or nonfootnoting, as the case may be? I’d welcome further opinions from everyone, but especially thoughts from those directly involved.

Related TeleBlog item: .Epub coming to Feedbooks: Merci, Hadrien!.


  1. I just tried “Through The Looking Glass” in FBReader and didn’t notice any slow paging, so the problem is probably in OpenBerg.

  2. dotReader supports .epub? The last release, from June, does not work on any of the .epub files I tried. So far as I know the list of released readers for .epub is still short: DE, OpenBerg, and FBReader. With FBReader providing limited support (no TOC or CSS).

    With all the FeedBooks titles now available, perhaps we can get early confirmation of other .epub compatible readers in the wild.

  3. Alan re .epub: Big thanks for reminding me of the wonderful FBReader’s capabilities as an .epub client! Ironically that’s the program I’m using to read public books on my Nokia 770. As for OSoft/dotReader, I’m gonna have the same ‘tude as toward Adobe or any other company. I want the world to know of any shortcomings in implementing the standard! Meanwhile I’m going to test out a file from Hadrien to see how it fares with FBReader. Thanks again. David

  4. I think Jon Noring was spot on with his comments about MS Reader, .epub, footnotes and popups. On an ebook reader, it makes more sense to have the footnote/endnote as a popup. This way, you don’t have the disadvantage of allways having it take up screen realestate, as well as the problems of pagination and keeping the footnote at the bottom. It makes perfect sense to me.

    Speaking of MS Reader, I have said it here before – MS Reader is the best ebook reader software (so far) in the way it handles the reading experience. It’s a shame that MS is so wrapped up in being proprietary that the software is so limited in other ways (platform, format, etc.). I wish either MS would extend Reader to support .epub and other platforms, or someone else would create a reader that works as nicely as Reader, but is more widely useable.

    The closest cross-platform, multi-format reader so far seems to be FBReader, which I use and love. However, compared to MS Reader, it has a ways to go in the “reader experience” department. MS Reader is especially nice on a Tablet PC, with a stylus.

    One problem with FBReader in particular is the way it handles paging. With some ebook formats, FBReader places the start of a new chapter at the top of the page, as it should. In some other formats, the chapters appear wherever they happen to be in the flow of text. For example, in HTML, chapter breaks work. In .epub and .oebzip, they don’t. This may seem like a minor issue, but it detracts from the reading experience.

    I guess what this long rant of mine sums up to is this – most of the problems with .epub that are being complained about are in the reader software (as noted by others). I wish I could do something about this situation myself, but I’m not a programmer.

  5. couple of mentions of “open source” in this post…

    “Open source creation tool available as well:”
    Where can the creation tool and source be found?

    “should thank the open-source-based Feedbooks site ”
    Where is the source for Feedbooks available?

    Links, please.

  6. FBReader should work fine with our files, although the lack of support for CSS and TOC means you won’t get the full experience.

    What I really like with FBReader is the fact that it supports hyphenation within the software.

    I’ll try soft hyphens in the future, to see if this is supported in DE.

  7. JRMurphy: Catch up with Feedbooks and let us know if you have problems getting what you need. I suspect Hadrien will do his best to be helpful whether or not his arrangement is formally open source. If I’m wrong on this, I’ll welcome clarification. Thanks.

    Hadrien: I’m hoping to test FBReader with Feedbooks in the next day or so. Meanwhlie continued congratulations and I hope you’ll do what you can to help JR and, if asked, Project Gutenberg.


  8. Well I can certainly help anyone trying to create epub files but the way Feedbooks generate these files won’t work on any of the other public domain books website.

    We don’t transform one type of file format to another format, we build these epub files from the ground up using our database.

    On Manybooks and PG, most of the books are just one huge chunk of text, while on Feedbooks, every book is divided into multiple elements in a database: part/chapter/section/text. That’s why we can’t simply copy/paste or use some kind of bot to add books on Feedbooks: during this process we’re also adding a semantic information. We use this information to create a TOC and more graphical elements, like our chapter header in PDF (or ePub if you open them in OpenBerg or DE).

    It’s pretty easy for a website like PG to support ePub: once you have an HTML file you can easily add the metadata and zip the whole thing. But with the way all of these other websites work, it’ll be pretty hard for them to get a TOC, footnotes and part/chapter headers.

  9. Hardien re .epub: Thanks for your observations. While it would appear that your particular setup isn’t transferable right now, you’ve done two things:

    1. You’ve said that some kind of .epub is possible from public domain sites even if it’s not as slick as at yours. There’s also the issue of speed. But I’ll be shocked if the appropriate open source tools don’t appear.

    2. More importantly, you’ve opened up a new issue–the question of whether PG and Manybooks.net and others might want to imitate your database approach. In a sense you’re applying a WordPress philosophy to the creation of .epubs, and even as a nonprogrammer, I find it interesting, given all the possibilities. How much of this is proprietary and how much would you be willing to share? I hope you can share. But it’ts an individual decision dependent on the business model, and in any event, you’ve made another point along the way: you don’t have to be Adobe to create .epub files.

    Meanwhile, just now, I did try to download books with .epub but saw some error messages. No big deal. You are, after all, in beta.

    At any rate, keep up the experimentation and keep giving us progress reports! Whether you’re little Feedbooks or S&S HarperCollins, the TeleBlog will be very supportive when people dare to try the new.


  10. About the message error: now that we’ve added epub files we have a lot more files in the cache, I need to change a few things, I really didn’t expect people to massively download epub this way. All the files already in cache should work and I’m working on fixing this problem right now.

    1. With ePub I don’t expect speed to be an issue.

    2. If we opened our source, it would be worth a lot of money for publishers, not much for the community. Our system is very complicated to replicate: it uses FreeBSD, Ruby on Rails, flex and a very complex production environment. We’re constantly improving the whole system, and in the future it’ll run on multiple dedicated machines (that’s what we need if we want to fully support the RSS generation and delivery in the future).
    I’d like to open Feedbooks but in two different ways:
    – You’re right about the blog approach: I’d like to make uploading a book easier than it currently is on Feedbooks. This way, people will be able to contribute to add semantic information to public domain books or if they’re publishing their own work, get an easy way to have their work published in multiple formats. This is aimed at non-programmer and shouls be the easiest way possible.
    – Have an open API. We already have an API for our “outputs”. The upcoming iLiad application is based on this and enable to search through the books on Feedbooks and download them, display favorites and similar books, or massively sync your RSS/newspaper subscription. These features can be added to anyone on any software. For example we could imagine a library in Second Life accessing all of our books this way, or an application on a smart-phone downloading books this way (instead of a mobile website, you get much more features with a dedicated application). But we’ll also work on an API for “inputs” in the future. This mean that any website could generate epub, PDF or any format supported by Feedbooks this way. Fan fiction websites such as http://www.fanfiction.net could be 100% e-books enabled that way. Or an amateur writer publishing his book under a Creative Commons license could have his books available as an ebook while he’s still publishing new chapters on his blog.

  11. Hadrien: Good to know you used open source tools to create your system. Sorry about the complexity. But maybe this is an argument for a consortium of public domain sites teaming up to provide a multiformat solution. What’s more, perhaps the consortium could make its services available to small publishers for little fees, voluntary or required, thereby generating a bit of an income stream. I like the idea of a database approach, if practical, because something broken into components can be endlessly manipulated just as in WordPress.

    Detail: I did download some files from Feedbooks and couldn’t get them readable in Adobe Digital Editions. Because of a glitch? Or are special steps necessary? Can you tell us what steps we should take for this to happen? Or are your .epub files readable in OpenBerg but not DE? If so, do you think the problem would be at your end or at DE’s end?

    As a booster of the idea of .epub, I don’t mind difficulties along the way–jut so they’re addressed. The more open .epub folks can be about their problems, the more likely a solution can be found through exchange of knowledge. Let the critics jeer. Once Feedbooks and other sites have .epub worked out, life will be a lot simpler for lovers of public domain works!


  12. I’ve tried dozens of files with DE and they all worked. But compared to the other readers, DE was VERY slow to load the files (dozens of seconds on War & Peace). Once the file is loaded and you’re in the main text, no speed issues anymore. These problems with DE are based on the fact that DE needs to paginate the whole book. Every time you resize or change the font size it takes a few seconds. If the file is very large, you have to wait much longer.

    Initial loading of the file is much faster if you add the file to your library and then open it, instead of directly opening the file with DE. Maybe there’s some preprocessing involved in this case ?

    Overall, if DE had hyphenation the look of the file would be almost like PDF: http://www.mobileread.com/forums/showpost.php?p=100180&postcount=2

    Which book have you tried with DE ?”

    Best regards,

The TeleRead community values your civil and thoughtful comments. We use a cache, so expect a delay. Problems? E-mail newteleread@gmail.com.