Citing Websites

[Hitch]

Originally Posted by DaleDe

There is a standard for page numbers in ePub 3 that can be very consistent and even match the hardcopy book.

@DaleDe: you're absolutely right, I forgot.

@NJ: See? What DaleDe said. ;-)

My bad, I completely spaced ePUB3.

Hitch

[Notjohn]

None of the ebooks I have had occasion to cite were equipped with page numbers.

I understood the problem here to be that of the academic writer citing book pages, not the author-publisher building a cite-able ebook.

Will that epub3 pagination carry over to the "mobi" that KDP makes of it? If so, maybe it's time for me to leave epub2 behind, tee hee.

[Turtle91]

In my unworthy opinion, the whole issue of the academics having a specific page number to cite is all about lazy professors. They don't want to spend ANY time confirming an author's source. The problem is that even pBooks have different versions and their page numbers change as well, so now the author has to include the version number in their citation... it's a never ending cycle imposed by professors onto the author (mostly students) because they can.... In my opinion the professor should be familiar enough with most of the main documents in the field they teach, or spend the 5 minutes to read the cited material and BECOME familiar with it... Then the references can simply be a chapter number or a subsection name, etc without having to cite a specific page.

For those non-professor types checking a citation - they are more than likely doing so to LEARN new info. They need to read the surrounding material anyway to get the proper perspective on what is being said. I've seen too many people quote only a sentence here or there and get the whole meaning wrong!

/rant

[Hitch]

Originally Posted by Notjohn

None of the ebooks I have had occasion to cite were equipped with page numbers.

I understood the problem here to be that of the academic writer citing book pages, not the author-publisher building a cite-able ebook.

Will that epub3 pagination carry over to the "mobi" that KDP makes of it? If so, maybe it's time for me to leave epub2 behind, tee hee.

Hmmm....dunno. I haven't tried that yet, to observe the result. It's an interesting experiment, though. I'll see if we have some space, in production this week, and ask one of my guys to try it.

Hitch

[Tex2002ans]

Originally Posted by Hitch

Hmmm....dunno. I haven't tried that yet, to observe the result. It's an interesting experiment, though. I'll see if we have some space, in production this week, and ask one of my guys to try it.

This was way back in 2016, but I doubt much has changed on this front. Way back in the "Real Page Numbers" thread, me and Doitsu PMed each other and discussed it in some more detail.

He linked me to his EPUB3 of "The Madman by Kahlil Gibran" with EPUB3 page numbers:

https://www.mobileread.com/forums/sh...d.php?t=264508

The EPUB3 used this form of code:

<span epub:type="pagebreak" id="page7" title="7"></span>

+ a page-list in the nav.xhtml:

Spoiler:

    <nav id="pagelist" epub:type="page-list">
      <ol>
        <li><a href="../Text/book.xhtml#page7">7</a></li>
        <li><a href="../Text/book.xhtml#page8">8</a></li>
        <li><a href="../Text/book.xhtml#page9">9</a></li>
        <li><a href="../Text/book.xhtml#page10">10</a></li>
        <li><a href="../Text/book.xhtml#page11">11</a></li>
[...]
        <li><a href="../Text/book.xhtml#page69">69</a></li>
        <li><a href="../Text/book.xhtml#page70">70</a></li>
      </ol>
    </nav>

Doitsu had noted this at the time:

Originally Posted by Doitsu

The downside of page number target references is that there aren't that many ePub apps that actually support real page numbers.
IMHO, this was best solved in iBooks and, AFAIK, the only free app that shows real page numbers is Azardi, however Azardi doesn't actually use the page list; it generates the page numbers from the page number target anchors.

This is how the page numbers displayed in iBooks on an iPhone 4:

Originally Posted by Notjohn

Will that epub3 pagination carry over to the "mobi" that KDP makes of it? If so, maybe it's time for me to leave epub2 behind, tee hee.

If you convert the EPUB->MOBI, this is what a physical Kindle looks like (note the "Page 11" in the bottom left corner):



Note: At least back in 2016, the "real page numbers" DO NOT SHOW UP in Kindle Previewer. Only showed on physical devices.

Side Note: Doitsu mentioned you can Go To "real page numbers" on Kindle:

Originally Posted by Doitsu

On Kindle touchscreen devices you can toggle the display between: location, page number, time left in chapter, time left in book and no indicator at all. (In my screenshot the location number is 22.)
Note that eInk Kindles will only display the first page number if a page contains two page breaks.

BTW, if a book contains "real page numbers," you can also enter a page number in the Go To dialog box. (In books without "real page numbers" no fake page numbers will displayed and the Go To Page option is disabled.)

Here is the picture he gave me of that:

Originally Posted by Notjohn

None of the ebooks I have had occasion to cite were equipped with page numbers.

I understood the problem here to be that of the academic writer citing book pages, not the author-publisher building a cite-able ebook.

I see them as intertwined problems. And the serious wrench occurs when the document is "born-digital" in the first place, and there ARE NO PAGE NUMBERS. Trying to force "pages" onto something like that is absurd and won't work, which is why you need to trend towards a more format-neutral citation method (chapters, subchapters).

And on the building side of things, inserting "real page numbers" is still a pain in the butt.

I discussed the three different pagination methods in detail below. EPUB2 has page-map [obsolete ADE-only] + pageList, and EPUB3 has page-list:

https://www.mobileread.com/forums/sh...02#post3404302

If you want some help generating the page anchors, there is Doitsu's helpful Sigil Plugin:

https://www.mobileread.com/forums/sh...d.php?t=265237

[Doitsu]

Switching to epub3 only for page numbers might not be such a good idea for Amazon KDP authors, because, AFAIK, KindleGen only supports NCX pageLists or ADE page-map xml files.

Also, last I checked, Amazon KDP didn't support "real page numbers." AFAIK, only big publishing houses who offer both print and Kindle books can offer Kindle books with "real page numbers" on Amazon.

Moreover, KindleGen apparently only supports the ePub3 items listed in the KindleGen 2.8 release notes, and nav page-list isn't listed.

[Hitch]

Originally Posted by Doitsu

Switching to epub3 only for page numbers might not be such a good idea for Amazon KDP authors, because, AFAIK, KindleGen only supports NCX pageLists or ADE page-map xml files.

Also, last I checked, Amazon KDP didn't support "real page numbers." AFAIK, only big publishing houses who offer both print and Kindle books can offer Kindle books with "real page numbers" on Amazon.

Moreover, KindleGen apparently only supports the ePub3 items listed in the KindleGen 2.8 release notes, and nav page-list isn't listed.

Hey, Doits:

Well, not in the routine course, I'd agree.

But, vis-a-vis what you said about "real page numbers," that's not totally correct--in our experience, I mean. Our experience--which is limited, granted, in this particular regard--is that if we submit a MOBI with RPNs (Real Page numbers acronym by yours truly) embedded into the coding, using IDs (e.g., "id="page7" or "p7" or whatever), and we link to those from a separate page index, or in the index, etc., we've seen Amazon enable the RPN functionality in that eBook. Without a page list or XML page-map.

Offered solely FWIW.

Spoiler:

(For that matter, remember THAT guy? The one that came to my firm first, and when I tried to explain the various issues to him, promptly came here to get a "real" answer? The one who TYPED his RPNS in, on the page, at the page breaks? THAT guy allegedly and apparently [based on some research I did] got RPN functionality enabled, with Amazon. Through simply manually entering those page numbers. So...there's that. I think that means that it's hard for us to say definitively what will, and won't, enable/turn on the RPN functionality within the Kindleverse.)

Hitch

[Tex2002ans]

Originally Posted by Doitsu

Switching to epub3 only for page numbers might not be such a good idea for Amazon KDP authors, because, AFAIK, KindleGen only supports NCX pageLists or ADE page-map xml files.

[...]

Moreover, KindleGen apparently only supports the ePub3 items listed in the KindleGen 2.8 release notes, and nav page-list isn't listed.

Good reference.

PS. I knew those PMs would come in handy some day. Now your genius finally sees the light of day! And I did steal your thunder, just like I said I would!

Originally Posted by Hitch

Spoiler:

(For that matter, remember THAT guy? The one that came to my firm first, and when I tried to explain the various issues to him, promptly came here to get a "real" answer? The one who TYPED his RPNS in, on the page, at the page breaks? THAT guy allegedly and apparently [based on some research I did] got RPN functionality enabled, with Amazon. Through simply manually entering those page numbers. So...there's that. I think that means that it's hard for us to say definitively what will, and won't, enable/turn on the RPN functionality within the Kindleverse.)

lol. Why yes, I linked to that glorious topic in my previous post!

[JSWolf]

Originally Posted by Hitch

Honestly, NJ, I kind of disagree that Location #'s are worthless. Why do you say that? At least they're consistent, unlike "page numbers" as created by ePUB devices, each to their own calculation.

Hitch

If the ePub page numbers are ADE type page numbers, then they are consistent. If not, then they are useless. So if you are reading the book in a device/app/program that doesn't use ADE page numbers, then before you cite, open the ePub in something that uses ADE page numbers and then cite.

[Tex2002ans]

Originally Posted by JSWolf

If the ePub page numbers are ADE type page numbers, then they are consistent. If not, then they are useless.

No. That's just plain wrong. Both ADE-based page numbers and Amazon's Location #s have the same flaws, just different scales.

I thoroughly discussed this in Post #166 and the surrounding posts in the "Sick of Amazon Kindle books without Page Numbers... " thread. I categorized them both as Byte Methods, and detailed the Pros/Cons of each.

I agree that ADE is a slightly better Byte Method though, because it is compressed... so it can handle redundant code a bit better, but I wouldn't rely on it for citations in any way.

Originally Posted by JSWolf

So if you are reading the book in a device/app/program that doesn't use ADE page numbers, then before you cite, open the ePub in something that uses ADE page numbers and then cite.

There is a vast ocean of ebooks + digital documents that are not in EPUB. And there are many very minor versions of ebooks out there (for example, I think I remember using my Game of Thrones EPUB as an example, where a version I purchased when it came out was a few ADE-pages off from the updated one I downloaded a few years later).

And depending on which tool you use to convert, the output can be wildly different, throwing off the Byte Method page numbers.

Imagine if a future Calibre or KindleGen conversion changed something very minor like <i> -> <em>, or <i> -> <i class="calibre123">. It wouldn't throw off the total bytes by much, but it will add up throughout the entire book.

Or take a Calibre conversion using in-line styling instead of a CSS Stylesheet. The text is the same, the display is the same, but the total bytes are way different.

What if I opened the EPUB and did a Prettifying on the code? The total bytes change. Imagine a hideously coded book (or a large enough one), and all of that whitespace will add up and throw off the Byte Method.

What if you open the book in Sigil, and it reorganizes the HTML files into subfolders. Links change:

<a href="Chap1.xhtml">Chapter 1</a>

becomes

<a href="../Text/Chap1.xhtml">Chapter 1</a>

Imagine something like an Index, with thousands of links going everywhere in the book, and that minor code change adds up. Same book, same functionality, same text... vastly different bytes.

[DaleDe]

Originally Posted by Tex2002ans

No. That's just plain wrong. Both ADE-based page numbers and Amazon's Location #s have the same flaws, just different scales.

I thoroughly discussed this in Post #166 and the surrounding posts in the "Sick of Amazon Kindle books without Page Numbers... " thread. I categorized them both as Byte Methods, and detailed the Pros/Cons of each.

I agree that ADE is a slightly better Byte Method though, because it is compressed... so it can handle redundant code a bit better, but I wouldn't rely on it for citations in any way.



There is a vast ocean of ebooks + digital documents that are not in EPUB. And there are many very minor versions of ebooks out there (for example, I think I remember using my Game of Thrones EPUB as an example, where a version I purchased when it came out was a few ADE-pages off from the updated one I downloaded a few years later).

And depending on which tool you use to convert, the output can be wildly different, throwing off the Byte Method page numbers.

Imagine if a future Calibre or KindleGen conversion changed something very minor like <i> -> <em>, or <i> -> <i class="calibre123">. It wouldn't throw off the total bytes by much, but it will add up throughout the entire book.

Or take a Calibre conversion using in-line styling instead of a CSS Stylesheet. The text is the same, the display is the same, but the total bytes are way different.

What if I opened the EPUB and did a Prettifying on the code? The total bytes change. Imagine a hideously coded book (or a large enough one), and all of that whitespace will add up and throw off the Byte Method.

What if you open the book in Sigil, and it reorganizes the HTML files into subfolders. Links change:

<a href="Chap1.xhtml">Chapter 1</a>

becomes

<a href="../Text/Chap1.xhtml">Chapter 1</a>

Imagine something like an Index, with thousands of links going everywhere in the book, and that minor code change adds up. Same book, same functionality, same text... vastly different bytes.

Your points are valid for a book that is constantly undergoing changes but such a book is not normally being used as a reference. One thing the byte method does help is that images being changed do not affect the page numbers since they are not counted, only the reference is counted. eBooks are a new world for referencing since they tend to be changed more often than a paper book.

Dale

[Hitch]

Originally Posted by DaleDe

Your points are valid for a book that is constantly undergoing changes but such a book is not normally being used as a reference. One thing the byte method does help is that images being changed do not affect the page numbers since they are not counted, only the reference is counted. eBooks are a new world for referencing since they tend to be changed more often than a paper book.

Dale

Dale:

Well....you're thinking of trade-pubbed "reference" books of non-fic. I can tell you that the non-fic that goes through here gets changed ALL. THE. TIME. I've got accounting books, for example, that we do annually to reflect changes in Tax law. Things like that.

Offered solely FWIW.

Hitch

[DaleDe]

Originally Posted by Hitch

Dale:

Well....you're thinking of trade-pubbed "reference" books of non-fic. I can tell you that the non-fic that goes through here gets changed ALL. THE. TIME. I've got accounting books, for example, that we do annually to reflect changes in Tax law. Things like that.

Offered solely FWIW.

Hitch

Sure, I understand this kind of book but when someone needs to reference it I would hope they can reference a particular edition since the data itself is likely to be wrong if they reference the wrong edition.

Dale

[JSWolf]

Originally Posted by Tex2002ans

And depending on which tool you use to convert, the output can be wildly different, throwing off the Byte Method page numbers.

I would not cite from a converted eBook. It has to be in a publisher released format. I would never buy an eBook from Amazon I might want to cite. The ADE page number would be off. It would have to be an untouched retail ePub.

[Tex2002ans]

Originally Posted by DaleDe

Sure, I understand this kind of book but when someone needs to reference it I would hope they can reference a particular edition since the data itself is likely to be wrong if they reference the wrong edition.

In many cases, the minor digital edition changes are not marked, so you would have no idea which ebook "version" you have exactly.

Example #1

Again, I go back to my example of Game of Thrones. It was an EPUB I downloaded and purchased from B&N when it came out. A few years later, I downloaded again, and believe I got the 15th EPUB revision (only because B&N appended "v15" to the filename).

Enough changed in the book, that the ADE-page numbers were thrown off between v1 and v15. They did things like changed lists to use <div>+CSS instead of <li>, typo corrections, fixed some italics, [...].

If I never had access to the v1 EPUB on the day of release, or I wasn't code-savvy, I wouldn't have a clue anything changed.

There is also typically no way to get OLDER revisions of the EPUBs..... I don't have any way to get my hands on any of the inbetween v1->v14 versions. Typically it just gets replaced with the latest and greatest.

And this is only B&N!

• Maybe 15 revisions on B&N were pushed.
• Maybe 5 on the publisher's website, because they only pushed the larger revisions.
• Maybe 20 on iBooks (adding in little iBooks-only CSS tweaks, etc., etc.).
• Maybe 30 on Kindle (code to get around poorly displaying stuff on pre-KF8 Kindles, or updating code when new higher-def Kindles were released).
  • And to solve a Kindle-only problem, there would be no need to revise the EPUBs.

Example #2

I mostly do work in Non-Fiction books. A large fraction of my work tends to be cleaning up poorly converted EPUB editions. It is the same book, the same text, but I just go in and strip out crap like:

<p class="paragraph"><span>This is a <span class="italics">sample</span> sentence.</span></p>

<p>This is a <i>sample</i> sentence.</p>

It is STILL the "First Edition" of the book. No text has changed, but there was serious HTML+CSS cleanup happening in the backend. Some publishers may mark something like "Second Ebook Edition, November 2017" to the copyright page, but most don't (nor could that be relied upon).

And again, what is large enough to count as a "change". If I fix a single typo? If I change a single line in the CSS? If I stripped out the useless <span>s?

I again go back to one of the largest EPUBs I ever worked on, a 1.4 million word beast (15 years of articles from a journal), the minor OCR+typo corrections threw off the ADE page numbers by a handful (and Kindle Locations would be thrown off even further).

Originally Posted by JSWolf

I would not cite from a converted eBook. It has to be in a publisher released format. [...] It would have to be an untouched retail ePub.

See Example #1 above. Even if it was "untouched", you have no idea which minor revision you have.

And an "untouched retail EPUB":

What if I have to strip the DRM? What if I purchased an EPUB from iBooks and I can't strip the DRM, so I can't get the EPUB into an ADE-based reader? That's touching the retail EPUB! (Not to mention, the ibooks EPUB might be different from the EPUB in other stores.)

What if the EPUB is being read on a non-RMSDK ereader? Does the page numbering on Kobo's KEPUBs follow ADE? What if I use a Readium-based reader?

What if, technically unsavvy user, tossed their untouched EPUB into Calibre. They then used Calibre to push to the Kobo. What if *GASP* Calibre helpfully did some tweaks to the EPUB for them, so it displays properly on their device. They would be none the wiser.

What if you use something like Kindle Unpack on your Amazon-purchased MOBI? That's altering the file!

And then look to the future, let's say 20 years from now. What happens when I purchase my ebook in XYZ format? XYZ format does not use ADE's page numbering scheme. Should I convert XYZ->EPUB using Calibre, because ADE is the one and only TRUE way to cite ebooks?

According to your own rules, you CANNOT alter the purchased ebook. Then how do you recommend citing the text in ebook format XYZ?

Side Note: Or if you don't like my abstract format XYZ, then pick an old format, like any of the books in the LRF + IMP + LIT section of MobileRead.

Originally Posted by JSWolf

I would never buy an eBook from Amazon I might want to cite. The ADE page number would be off.

Millions of other people purchase and read non-EPUB ebooks, and millions of people read in non-RMSDK/ADE readers. You can't just try to force JSWolf's personal preferences down everyone else's throats.

More broad, format-neutral citation is superior, because it works across all the different formats (Hardcover, Large Print, PDF, HTML, EPUB, MOBI, XYZ, [...]).