Fractional Page Numbering

[rcentros]

Originally Posted by MGlitch

I agree the conversion from page numbers to word counts is much the same as metric to imperial. There’s no real good reason not to convert, we simply don’t because we’re all accustom to the system we use.

Or, on the other hand, there's no real good reason to convert. People in the United States, for example, understand miles and Fahrenheit. Who cares if it matches the rest of the world? If the standard unit of measure was an average jack rabbit hop, times 221.7, and everyone was used to that, why should they change to some other standard? What's wrong with using what we're accustomed to? Why should we change for the sake of changing?

[rcentros]

Originally Posted by ZodWallop

So when a hard cover is published as a mass-market paperback where the page numbers change due to the size of the font used, does that mean the paperback doesn't have actual pages? ...

The whole point of this thread is that the OP wants to develop a page numbering system that will remain consistent between a specific edition of a paper book and its eBook counterpart. I realize this thread has been hijacked (not that I haven't been part of that) for yet another debate on page numbering. But, personally, I think the original aim here would be great for students who are working with a specific edition of a textbook. (Personally I would like this for novels also — but I think that's beside the point.)

[Tex2002ans]

Originally Posted by j.p.s

All I have done is point out that existing devices will display, in addition to roman numerals and ascii integers, page numbers such as 4.5 (and I assume 13b).

Okay, 13b I can see some use for.

Good that you were testing that. I never really messed with much beyond your usual Roman Numerals + basic numbers.

Originally Posted by j.p.s

I have not tried to predict characters per screen or anything like that, just inserting anchors with fractional page ids in between the existing anchors with integer page ids, [...]

Well, initially you said:

"and still show change for each ebook page turn."

This made me think you were trying to emulate some sort of "Screens", but hardcoding even more of them. So you would arbitrarily split every page into .1, .2, .3, to try to get closer to a "screen per page".

Still would recommend against that because it would interfere with blind readers who want to skip to next pages.

Originally Posted by MGlitch

Of course going by word count has the benefit of not being subject to the medium. I’m sure there are drawbacks such as longer works getting unwieldy in terms of the numbers. But that can be worked around.

There's also the case of what's a "word".

I discussed this all in the linked threads, and came up with the concept of "Format-Specific" and "Format-Neutral".

A "Format-Specific" example would be:

Page Number (like in a physical book). Only fits with that specific page/font size, those specific margins, [...].

A "Format-Neutral" example would be:

Paragraph Count. It doesn't matter if it's physical, Large Print, HTML, Ebook, the amount of paragraphs should be the same across formats. (Also think laws: "see Section 8, paragraph 3".)

Every reference format is going to have pros/cons. Here's what I wrote on Word Counts:

• Word Count
  
  • Note: Depends heavily on what you count as a "word".
    
    • Do you count numbers as a word? There are quite a few differences between "word counts". For example, Microsoft Word/Sigil/Calibre all disagree and vary in "word count" by tiny margins.
  • My personal thoughts on this method... I believe it is "too specific", and would lead to too many errors if trying to reference it.
    

Elsewhere in the thread, I also discussed a few more complicated edge cases with Word Counts. Here's some newer examples:

• Generated text (Javascript)
• Tables
  
  • Does each cell count as 1 word? Or do Tables not count?
    
  • What about a 5x5 table of numbers? 25 words?
    
• Text that displays in one format and not another.
  
  • Think MOBI vs. KF8. MOBI might display an image of a complicated table, KF8 displays an actual <table>.
• Math
  
  • Is "x + y = 2" 3 words or 0? What about "x+y=2"? 1 word?
    
  • 1½ (3 words? 2? 1?)
    
  • 1 1/2 (3 words? 2?)
    
• Languages without spaces
  
  • Many Asian languages don't use spaces at all. So what's a "word"? Each symbol?
• Spaces
  
  • How do you deal with characters like zero-width joiners, or no-break spaces. Does that mean you treat both sides as a "single word"?
    
• Dashes/Hyphens
  
  • (I grabbed a few of these from Wikipedia's En Dashes)
    
  • The hospital–nursing home connection (5 words? 4?)
    
  • The pro-conscription–anti-conscription debate (4 words? 3?)
    
  • love–hate relationship (3 words? 2?)
    
  • love-hate relationship (3 words? 2?)
    
  • non–Euclidean geometry (3 words? 2?)
    
  • See pages 123–125. (4 words? 3?)
    
• Periods/Colons
  
  • 11:00 a.m.–1:00 p.m. (4 words? 3? 7? 8?)
    
  • 11:00a.m.–1:00p.m. (1 word? 2? 5? Other?)
    
  • U.S.A. (1 word? 3?)
    
  • As H.M. Smith said. (5 words? 4?)
    
  • As H. M. Smith said. (5 words?)
    
  • "I…I don't know." (4 words? 3?)
    
  • "I...I don't know." (4 words? 3?)
    
  • "I-I don't know." (4 words? 3?)
    

If anything, I would say Paragraph Count is more neutral/steady across mediums. It's pretty well-proven in cases like laws ("see Section 8, paragraph 3, clause a"). And laws have to be written in a way and published in many different formats, they can't rely on something too specific/variable.

[MGlitch]

Originally Posted by rcentros

Or, on the other hand, there's no real good reason to convert. People in the United States, for example, understand miles and Fahrenheit. Who cares if it matches the rest of the world? If the standard unit of measure was an average jack rabbit hop, times 221.7, and everyone was used to that, why should they change to some other standard? What's wrong with using what we're accustomed to? Why should we change for the sake of changing?

Accuracy? There's a reason metric is used in the hard sciences even in the US, to get everyone on the same page, if you'll pardon the pun.

The aim on the thread was to create a system which is consistent, but it's doing that by establishing a false base as the setting (a physical book of a chosen edition), and inflicting the pages there to ebooks. Which of course falls apart if there's a different edition of the book in physical form say as a large print edition.

Meanwhile 500 words into a 'regular print', ebook, and large print of the same book will get you to the same spot regardless of typography, or page size.

[Tex2002ans]

Originally Posted by MGlitch

Meanwhile 500 words into a 'regular print', ebook, and large print of the same book will get you to the same spot regardless of typography, or page size.

Incorrect. Close, but incorrect. (Look at the edge cases above.) Word Counts can vary wildly, especially in the case of ebooks.

For your typical Novel, a raw word count would probably be "okay", but anything more complicated than that (Non-Fiction, Math, Javascript-enabled books)...

And the larger the books, the more different word count algorithms diverge.

[MGlitch]

Originally Posted by Tex2002ans

Incorrect. Close, but incorrect. (Look at the edge cases above.) Word Counts can vary wildly, especially in the case of ebooks.

For your typical Novel, a raw word count would probably be "okay", but anything more complicated than that (Non-Fiction, Math, Javascript-enabled books)...

And the larger the books, the more different word count algorithms diverge.

All of which are much more simply resolved than what constitutes a page. Especially as we move away from physical pages and into digital books where pages don’t actually exist, meanwhile words still do.

[ZodWallop]

Originally Posted by rcentros

I realize this thread has been hijacked (not that I haven't been part of that)...

yeah, me too. For something that, in the end, I'm not too fussed about. What can I say, it's a slow day at the office

[Uncle Robin]

Originally Posted by Tex2002ans

Incorrect. Close, but incorrect. (Look at the edge cases above.) Word Counts can vary wildly, especially in the case of ebooks.

For your typical Novel, a raw word count would probably be "okay", but anything more complicated than that (Non-Fiction, Math, Javascript-enabled books)...

And the larger the books, the more different word count algorithms diverge.

(e.a.)

Now I'm curious: This thread has been about "real" page numbers, arbitrarily assigned from a specific edition of a paper copy of a book. How many paper books are Javascript-enabled?

[Tex2002ans]

Originally Posted by MGlitch

All of which are much more simply resolved than what constitutes a page. Especially as we move away from physical pages and into digital books where pages don’t actually exist, meanwhile words still do.

Forgot to discuss Footnotes too.

So let's say you wanted a Word Count as a reference to a specific location in a book... where do Footnotes belong?

You could code them at the bottom of the file:

<p>This is a paragraph.[1]</p>

<p>This is a second paragraph.</p>

[...]

<hr />

<p>[1] This is footnote 1.</p>

You could have them within text in an EPUB3 <aside>:

<p>This is a paragraph.[1]</p>

<aside><p>[1] This is footnote 1.</p></aside>

<p>This is a second paragraph.</p>

or you could have Endnotes:

<p>This is a paragraph.[1]</p>

<p>This is a second paragraph.</p>

* * * (Completely separate file) * * *

<p>[1] This is endnote 1.</p>

Let's say you put your cursor in the second paragraph (or the footnote!)... each one of these will give you wildly different results.

Should footnotes even count towards word count? They're arguably not a part of the main text... and may not even be displayed depending on which format you're reading in.

A more "neutral" scheme (like Paragraphs) wouldn't be as wildly variable. (*Although it runs into other hard edge cases... like this footnotes issue.)

[MGlitch]

Again still an easier thing to resolve than page numbers that don’t exist. It’s not like citation methods aren’t well documented with various styles as fits the disciplines of the book.

I’ll save you the trouble of finding other cases, anything which deals with something hard and finite is going to be easier than dealing with an arbitrary system being fixed into another arbitrary system to try and make a third arbitrary system in the name of accuracy.

[JSWolf]

Originally Posted by MGlitch

I agree the conversion from page numbers to word counts is much the same as metric to imperial. There’s no real good reason not to convert, we simply don’t because we’re all accustom to the system we use.

Of course going by word count has the benefit of not being subject to the medium. I’m sure there are drawbacks such as longer works getting unwieldy in terms of the numbers. But that can be worked around.

Not that I want to get into the debate over systems used to measure progress in ebooks again.

Word count is not going to cut it. Not many would like to see page numbers that high. Kobo used to use 340 words = 1 page for KePub until the 4.18.x firmware where it's not 1 screen = 1 page.

I really dislike 1 screen = 1 page for the full book. I don't mind for chapter page numbers. Word count used as page numbers is also a no way. The ending page number would be too high and that won't work for a lot of people.

We have ADE page numbers. Why not just go with that as it works well?

[JSWolf]

Originally Posted by leebase

The problem is "page" isn't that useful a concept when readers of the same book might be reading on a small phone, a large phone, an 6" eInk reader, a 13" eInk reader, or 7-12" tablets.

And...what size has the user/reader app set the font?

And this is why ADE page numbers work. Because they work regardless of all the issues mentioned.

[jhowell]

Originally Posted by rem736

amazon's location method counts each line of text. so it's really the chapter and verse method without the chapter.

That is not true. Amazon "locations" each count 150 bytes of HTML3 data in the MOBI7 format as produced by Kindlegen. The same text with different styling will have different location numbers.

[JSWolf]

Originally Posted by ZodWallop

Okay, but for all that people say that, I have seen no proof that Adobe page numbers actually do correlate to a specific paper version of a book.

ADE page numbers are not meant to correspond to some pBook version. ADE page numbers are meant to be consistent no matter the screen size, font, font size, margins, line height, font size, etc. And they are. That's why ADE page numbers work and is the best page number system for eBooks.

[JSWolf]

Originally Posted by MGlitch

Meanwhile 500 words into a 'regular print', ebook, and large print of the same book will get you to the same spot regardless of typography, or page size.

Sorry, but no. Not going to work. The regular pBook and the large print pBook are not going to be in the same place 500 words in. You have different text before you get to the start of the story (in the case of a novel) and that right there throws off the count between the two versions.