Loading ...
Sorry, an error occurred while loading the content.

Re: [ebook-community] IDPF and multipart/related

Expand Messages
  • Lee Passey
    ... This is a perfect example of how personal biases effect design decisions. Apparently you are comfortable reading e-books in a web browser while sitting at
    Message 1 of 18 , Aug 1, 2007
    • 0 Attachment
      Bill Janssen wrote:

      >> Unfortunately, IE and Opera are the /only/ browsers that support
      >> parsing=20 and displaying MHTML.
      >
      > And 80% of the world's browsers, to start with, is bad?

      This is a perfect example of how personal biases effect design
      decisions. Apparently you are comfortable reading e-books in a web
      browser while sitting at a desktop computer, and therefore believe that
      everyone would be comfortable doing so. To me, reading books in a web
      browser sucks, and reading while sitting at a desk sucks even more. I
      think just about every journalist who has commented negatively about
      e-books has said that they don't think e-books will catch on because
      they can't imagine someone sitting at a computer to read for enjoyment.
      They don't get it, because they don't understand that reading e-books
      (at least non-PDF e-books) /doesn't/ mean sitting at a computer to read,
      and the e-book community at large hasn't done a very good job of
      communicating this fact.

      You believe that supporting IE and Opera alone gives 80% market coverage
      because 80% of the market has access to IE. I believe that supporting IE
      alone provides only 20% market coverage because only 20% of the market
      would be /willing to use/ IE as an e-book reading tool.

      Your argument is remarkably similar to that used by Michael Hart in
      favor of his degraded text format: simple ASCII provides 100% market
      coverage if only consumers would just lower their standards enough.

      >> I know of /no/ tools that exist to create these files. Even the IE
      >> function saves all binary data (such as images) as uuencoded (text) data.
      >
      > The fact that you don't know about them isn't an interesting point :-).
      >
      > As I remember pointing out to you before, almost all HTML-formatted
      > email is multipart/related. Every email-handling library contains
      > support for it. That's a lot of tools.

      No, that's a lot of programs which do something similar to that which
      would be required to create and present MIME-based e-books, although
      they do contain a fair amount of source code that could be used to
      create the necessary tools. (Of course, if the source code has
      restrictive licensing, such as the GPL, it would be pretty much unusable
      by any commercial enterprise.) You are suggesting that there are a lot
      of resources for programmers; you have not identified any existing tools
      for producers and consumers.

      This is one of the reasons I really ought to be supporting your
      MIME-based encapsulation proposal. I'm a programmer well-versed in
      Internet protocols and fairly knowledgeable about e-books. Writing some
      of the currently non-existent tools required to support a MIME-based
      encapsulation scheme wouldn't be difficult for me, so if a MIME-based
      format were accepted as a standard I could probably convince one of the
      major players to hire me at a premium. It wouldn't be good for consumers
      or for producers, but I ought to be able to take advantage of it personally.

      Now if the tools to create and manipulate a MIME-based encapsulation
      format really do exist, you ought to be able to write a brief set of
      instructions telling us how to create an e-book in that format much as
      John Rivlin has done for EPUB
      (http://www.ebooktechnologies.com/OCFUsingWinZip/OCFUsingWinZip.htm).
      Can you do so? If so, please share it with us.

      >> From my biased
      >> perspective, being able to read an e-book in IE, obtained directly from
      >> the web without being stored locally has virtually no value.
      >
      > That is indeed the problem. Groups like the IDPF are dominated by
      > folks biased to the old failed last-century model of the ebook, and
      > find it hard to accept that that future didn't, in fact, happen -- and
      > won't. Again, the way that you put that seems to me a demonstration
      > that you don't quite understand what's happening to publishing and
      > communications technology.

      Understand that I am not associated with the IDPF or any e-book
      publisher or software producer. My interests in e-books are as an avid
      reader and technophile; in other words, I am a consumer and nothing
      else. As a consumer, I don't believe in the future you are trying to
      promote. (When I say "don't believe" I don't mean that I have no faith
      in it, nor that I wish that it will not happen; I mean that rational
      extrapolation from consumer expectations and technological trends leads
      me to conclude that it /will not/ happen).

      I want e-books I can read on a device small enough to hold in one hand.
      I want e-books whose presentation I can alter easily to accommodate my
      failing eyesight. And I want an e-book I can read and enjoy while
      sitting on a houseboat in the middle of Lake Powell. I don't think that
      the e-book model prevalent at the beginning of this century has failed,
      and I don't think that technology is moving in a direction which will
      invalidate that model. So far, it seems to me that those who /do/
      believe this are a minority of one.

      >> So, if you value the option of being able to read an e-book in IE or
      >> Opera only, on a desktop computer, while connected to the Internet with
      >> a fat pipe, without regard to the difficulty of creating content ...
      >
      > But of course, that isn't the real situation. That's the situation as
      > you perceive it to be, much as it was 10 years ago, when the old ebook
      > ideas the IDPF is, quite rightly, in existence to protect and preserve,
      > originated.

      No, it /is/ the real situation, based upon my experience as a consumer.
      Whatever your prognostications, you cannot dictate what my needs and
      desires should be. While it is possible that you are right and everyone
      else in the world is wrong, that's a position I tend to approach with a
      great deal of skepticism.

      >> I believe that the existence of an IDPF container format is probably due
      >> almost exclusively to Mr. Noring's advocacy of the OpenReader format.
      >> The IDPF members recognized the need, and fearing loss of control over
      >> the format, decided to move forward with it fairly expeditiously.
      >
      > I agree. Now, was that a good thing?

      Those people who agree with what he was proposing seem to think so. You
      think your proposal of using MIME encapsulation is superior. I'm
      inviting you to do the same thing Mr. Noring did. Flesh out your
      proposal, create a proof of concept, publicize it and let's see how
      widely it is adopted. People are not stupid, not even those who disagree
      with you. If it truly is better we'll find out soon enough.

      --
      Nothing of significance below this line.
    • Bill Janssen
      Lee, I don t think we re going to make much progress by arguing with each other. But I do invite you to review the OpenReader archives on this topic. Bill
      Message 2 of 18 , Aug 1, 2007
      • 0 Attachment
        Lee,

        I don't think we're going to make much progress by arguing with each
        other. But I do invite you to review the OpenReader archives on this
        topic.

        Bill
      • Jon Noring
        ... The OpenReader archives Bill refers to are found at: http://tech.groups.yahoo.com/group/openreader-format/messages Only part of the discussion focuses on
        Message 3 of 18 , Aug 1, 2007
        • 0 Attachment
          Bill Janssen wrote:

          > I don't think we're going to make much progress by arguing with each
          > other. But I do invite you to review the OpenReader archives on this
          > topic.

          The OpenReader archives Bill refers to are found at:

          http://tech.groups.yahoo.com/group/openreader-format/messages


          Only part of the discussion focuses on the container format.

          Jon Noring
          OpenReader Consortium
        • David Teller
          ... Well, that guide is hardly complete. There s nothing on how to create OPF files, which are way beyond the comprehension of mere mortals. Anyway, there s a
          Message 4 of 18 , Aug 2, 2007
          • 0 Attachment
            On Wed, 2007-08-01 at 12:26 -0600, Lee Passey wrote:
            > Now if the tools to create and manipulate a MIME-based encapsulation
            > format really do exist, you ought to be able to write a brief set of
            > instructions telling us how to create an e-book in that format much
            > as
            > John Rivlin has done for EPUB
            > (http://www.ebooktechnologies.com/OCFUsingWinZip/OCFUsingWinZip.htm).
            > Can you do so? If so, please share it with us.

            Well, that guide is hardly complete. There's nothing on how to create
            OPF files, which are way beyond the comprehension of mere mortals.

            Anyway, there's a Firefox extension for manipulating multipart/related,
            it's called MAF. And it shouldn't be too hard to write a simple Java
            applet to do the same.

            >
          • Nick Bogaty
            David, I certainly agree with your comment, well, that guide is hardly complete. There s nothing on how to create OPF files, which are way beyond the
            Message 5 of 18 , Aug 3, 2007
            • 0 Attachment
              David,

              I certainly agree with your comment, "well, that guide is hardly complete.
              There's nothing on how to create OPF files, which are way beyond the
              comprehension of mere mortals." While we are now seeing many large trade
              and academic publishers working with their conversion houses to upgrade from
              OEB to .epub (OCF/OPS), I understand that this is far beyond the resources
              that most "mere mortals" would have access to. It is my hope, and I'm sure
              the hope of IDPF members, that companies will leverage .epub to create
              consumer-grade authoring tools to simply "hit a button" to create a .epub
              file. We are beginning to see this with the "export as .epub" feature in
              Adobe inDesign CS3 and I'm sure we'll see similar authoring products from
              other companies in the near future. The IDPF also plans to help. We are
              currently working on releasing validation tools for .epub so companies and
              individuals can be sure that their .epub is really .epub.

              We're still quite early here (OPS 2.0 isn't even officially approved
              yet...early Sept. for voting), but it is our intention to make the creation,
              distribution and consumption of .epub files as easy as possible.

              -Nick
              --
              Nick Bogaty
              Executive Director
              International Digital Publishing Forum (IDPF)
              nbogaty@...
              www.idpf.org
              (212) 924-9081 direct
              (212) 208-0978 fax



              [Non-text portions of this message have been removed]
            • Bill Janssen
              ... I haven t looked too hard at doing it in Java (it s trivially easy to write a Python program to build MHTML bundles, and at one point I believe I posted
              Message 6 of 18 , Aug 3, 2007
              • 0 Attachment
                > Anyway, there's a Firefox extension for manipulating multipart/related,
                > it's called MAF. And it shouldn't be too hard to write a simple Java
                > applet to do the same.

                I haven't looked too hard at doing it in Java (it's trivially easy to
                write a Python program to build MHTML bundles, and at one point I
                believe I posted such a "tool" to the OpenReader mailing list), but a
                good starting place would be javax.mail.internet.MultiPart (see
                http://java.sun.com/products/javamail/javadocs/javax/mail/internet/MimeMultipart.html,
                part of http://java.sun.com/products/javamail/downloads/index.html).
                The JavaMail (and the JAF) code is open source under the CDDL
                (https://glassfish.dev.java.net/javaee5/mail/).

                From a cursory scan of the ChangeLog (the fact that I don't see the
                word "related" there :-), it looks to me that you'd still need to add
                support for the "related" subtype of "multipart", which basically
                involves adding some code to support local resolution of the "cid" URL
                scheme from "Content-Location" and "Content-ID" headers in the part
                headers. It would be a nice Java class exercise :-).

                Bill
              • David Teller
                In the mean-time, as there was no answer to my post regarding construction of .opf publications, I ll assume that my OpenBerg Rector is the only such tool. Or
                Message 7 of 18 , Aug 3, 2007
                • 0 Attachment
                  In the mean-time, as there was no answer to my post regarding
                  construction of .opf publications, I'll assume that my OpenBerg Rector
                  is the only such tool. Or at least the only such open-source tool. Not
                  that it's available for general use yet, but it's already usable by mere
                  mortals.

                  Well, too bad. I hoped I could draw some inspiration on usability from
                  other developers.

                  Cheers,
                  David,
                  Who often feels alone on the open-source e-Book front.

                  On Fri, 2007-08-03 at 10:10 -0400, Nick Bogaty wrote:
                  > David,
                  >
                  > I certainly agree with your comment, "well, that guide is hardly
                  > complete.
                  > There's nothing on how to create OPF files, which are way beyond the
                  > comprehension of mere mortals."
                  >
                • Lee Passey
                  ... 1. Actually, the creation of OPF files is almost trivial, and well within the comprehension of an average clerk/typist. Anyone who claims it is beyond the
                  Message 8 of 18 , Aug 3, 2007
                  • 0 Attachment
                    David Teller wrote:
                    >
                    > On Wed, 2007-08-01 at 12:26 -0600, Lee Passey wrote:
                    >
                    >> Now if the tools to create and manipulate a MIME-based encapsulation
                    >> format really do exist, you ought to be able to write a brief set of
                    >> instructions telling us how to create an e-book in that format much
                    >> as John Rivlin has done for EPUB
                    >> (http://www.ebooktec hnologies. com/OCFUsingWinZ ip/OCFUsingWinZip.htm.
                    >> Can you do so? If so, please share it with us.
                    >
                    > Well, that guide is hardly complete. There's nothing on how to create
                    > OPF files, which are way beyond the comprehension of mere mortals.

                    1. Actually, the creation of OPF files is almost trivial, and well
                    within the comprehension of an average clerk/typist. Anyone who claims
                    it is "beyond the comprehension of mere mortals" is trying to sell you
                    something.

                    About three years ago I started writing a simple Java Swing application
                    that would put a GUI around the OPF creation process. I set the project
                    aside when I realized that if you know enough to decide what data to put
                    in which the text fields you know enough to build the OPF file using a
                    simple text editor. After you've done about two of them you'll be saying
                    to yourself, "This isn't hard at all."

                    On the other hand, if a GUI tools is really required, the Mobipocket
                    Publisher and Mobipocket Creator create OEB publications using an OPF
                    file. Overdrive's ReaderWorks may do so as well but I'm not certain
                    about that. At the very least you could use ReaderWorks to create a
                    Microsoft Reader file and then use ConvertLit to extract a complete OEB
                    publication including the OPF file. (Actually, any tool that creates
                    Microsoft Reader files could be used in the same way).

                    An extremely comprehensive, yet readable, guide to everything you could
                    ever want to know about creating OPF files can be found at
                    http://www.microsoft.com/reader/developers/downloads/source.mspx

                    2. Even if one can't figure out how to build how to build an OEB
                    publication, that problem is pretty much irrelevant in the discussion of
                    MIME vs. ZIP. Assuming that I have prepared a valid OEB publication, Mr.
                    Rivlin's instructions tell me how to create a valid OCF file (aka EPUB,
                    for the recommended file extension). Assuming that I have prepared a
                    valid OEP publication, can anyone tell me how I can make a
                    multipart/MIME file without programming?


                    > Anyway, there's a Firefox extension for manipulating multipart/related,
                    > it's called MAF. And it shouldn't be too hard to write a simple Java
                    > applet to do the same.

                    I find the applet idea intriguing. Of course, the motivation for being
                    able to read e-books in a browser is to take advantage of the browser's
                    capability to 1. obtain resources from a network instead of a local file
                    system, and 2. render the marked-up content. I was not aware that
                    applets could interact with the host browser that way, instructing the
                    browser to download and display a given resource. Is this in fact
                    possible? Could you point me to someplace on the web where sample source
                    code might be found?


                    --
                    Nothing of significance below this line.
                  • David Teller
                    ... No, I m just stating the obvious: XML has never been meant to be human-readable/human-writable except in desperate situations. So, while typing an OPF file
                    Message 9 of 18 , Aug 4, 2007
                    • 0 Attachment
                      On Fri, 2007-08-03 at 14:04 -0600, Lee Passey wrote:
                      > 1. Actually, the creation of OPF files is almost trivial, and well
                      > within the comprehension of an average clerk/typist. Anyone who
                      > claims
                      > it is "beyond the comprehension of mere mortals" is trying to sell
                      > you
                      > something.

                      No, I'm just stating the obvious: XML has never been meant to be
                      human-readable/human-writable except in desperate situations. So, while
                      typing an OPF file is simple enough, as long as it doesn't involve
                      *understanding* anything, a task that requires actual *comprehension*,
                      such as finding the error in an OPF file, is beyond the comprehension of
                      my students in third year of computer science/engineering.


                      > About three years ago I started writing a simple Java Swing
                      > application
                      > that would put a GUI around the OPF creation process.

                      Yep, done that a few weeks ago. Saves even me a number of errors,
                      especially when you start having books with 100+ resources. Right now,
                      I'm trying to turn it into something professional-looking.

                      > After you've done about two of them you'll be saying
                      > to yourself, "This isn't hard at all."

                      That's a usual mistake made by people with a good understanding of the
                      basics. Yeah, it's probably not hard at all. Except that for most
                      people, it is.

                      > On the other hand, if a GUI tools is really required, the Mobipocket
                      > Publisher and Mobipocket Creator create OEB publications using an OPF
                      > file. Overdrive's ReaderWorks may do so as well but I'm not certain
                      > about that. At the very least you could use ReaderWorks to create a
                      > Microsoft Reader file and then use ConvertLit to extract a complete
                      > OEB
                      > publication including the OPF file. (Actually, any tool that creates
                      > Microsoft Reader files could be used in the same way).

                      And this is expected to be simple ? Understanding how to *compile*
                      ConvertLit took me about 1/2h, and I'm an experienced developer.

                      > An extremely comprehensive, yet readable, guide to everything you
                      > could
                      > ever want to know about creating OPF files can be found at
                      > http://www.microsoft.com/reader/developers/downloads/source.mspx

                      Thanks, I'll read it.

                      > 2. Even if one can't figure out how to build how to build an OEB
                      > publication, that problem is pretty much irrelevant in the discussion
                      > of
                      > MIME vs. ZIP.

                      That's true.

                      > > Anyway, there's a Firefox extension for manipulating
                      > multipart/related,
                      > > it's called MAF. And it shouldn't be too hard to write a simple Java
                      > > applet to do the same.
                      >
                      > I find the applet idea intriguing.

                      Actually, I meant for building books.

                      > Could you point me to someplace on the web where sample source
                      > code might be found?

                      >From the top of my mind,
                      http://java.sun.com/products/plugin/1.3/docs/jsobject.html

                      Cheers,
                      David
                      >
                    Your message has been successfully submitted and would be delivered to recipients shortly.