The Technology Vent and Rant Thread

[John F]

I still like the way pbook pages were "counted": the story itself was what was counted, not the the other stuff (cover, copyright, acknowledgments, ...).

[DNSB]

Originally Posted by madelonw

The Count Pages plugin doesn't count pages as they are numbered. In fact, the number it puts out is usually higher than actual print pages. I have my own columns for Fiction/Non-Fiction, and Category (i.e., novel, novella, short story, etc). I use the count pages result to say this is a novel, this is a novella, etc. It also has been useful in the past for selecting something to read that I will be able to finish during a 2-hour plane trip. Bottom line, the number is relative but I find it useful.

Which print pages? Unless the ebook is created from the same file as the print version, the page counts are very unlikely to be the same. I have several books in multiple print editions, one example being Eric Flint's Mother of Demons for which I have a hard cover, a trade paperback and a mass-market paperback. All 3 have different page counts. Which one is the correct one to claim as the "ONE TRUE PAGE NUMBER™" source?

OTOH, I also find the output from Count Pages to be useful as an indication of the book reading time.

OYAH, if you really want to confuse Count Pages, try feeding it a ebook that used Base64 to encode images as text. I had one ebook that had multiple Base64 encoded images which resulted in a rather large file and that Count Pages reported as having ~2500 pages. Extracting the images, converting them to JPEG and re-inserting them resulted in a much smaller ebook file and with ~250 pages.

[Quoth]

Originally Posted by John F

I still like the way pbook pages were "counted": the story itself was what was counted, not the the other stuff (cover, copyright, acknowledgments, ...).

Yes, and that is the same for a word count of any book. You only include the body.

Indeed sane paper books only number the body text pages from 1 on the first page of body.

[Quoth]

Originally Posted by DNSB

…
one example being Eric Flint's Mother of Demons for which I have a hard cover, a trade paperback and a mass-market paperback. All 3 have different page counts. Which one is the correct one to claim as the "ONE TRUE PAGE NUMBER™" source?

Exactly. There is no true page number except for text books in one edition. I remember one lecturer that specified an edition. It was the only text book I bought, being broke. He would start by "Now turn to page <whatever>". There was an attendance requirement or you couldn't sit any exam. So I had to go to his almost pointless lectures.

Though in order of preference if I had to choose:
mass-market (might not exist)
trade (more likely to exist at ebook time)
Hard back (comes out first, but ebook may not happen till Trade edition and not all titles get a hard back)

Large Print editions are getting rarer, and there would be regular edition.

Just deciding a fixed number of words (say 1200 per page and it's a word processor setting in LO) or characters (maybe 7,200 per page) is OK for novels, except for the reference work /text book and it is going to use a table based on a real edition. Adobe, while generally annoying, indeed support both those scenarios for years.

[Quoth]

Originally Posted by DNSB

OYAH, if you really want to confuse Count Pages, try feeding it a ebook that used Base64 to encode images as text. I had one ebook that had multiple Base64 encoded images which resulted in a rather large file and that Count Pages reported as having ~2500 pages. Extracting the images, converting them to JPEG and re-inserting them resulted in a much smaller ebook file and with ~250 pages.

Why did they do that? I thought that sort of image encoding was for ancient email systems.

[ratinox]

Originally Posted by Quoth

Why did they do that? I thought that sort of image encoding was for ancient email systems.

If by ancient you mean anything you sent or read today? Yup. Base64 is very much actively used for all sorts of binary blobs from embedded images to SPF/DKIM/DMARC data in the headers.

[DNSB]

Originally Posted by Quoth

Why did they do that? I thought that sort of image encoding was for ancient email systems.

Ghod alone knows why. The only reason I noticed was the massive page count since the encoded images were counted as text. I knew you could Base64 encode images and they were permitted in ePub but this was the first and only time I saw it done.

[Solitaire1]

Originally Posted by Karellen

Yep, just use Word Count. There cannot be a more reliable counter for the size of a book. 84,000 words is 84,000 words. No need for silly calculations to covert to a meaningless page count all so you can satisfy yourself of the size of a book.

Over the last few years, as I realised how useless a "page count" really is for ebooks, I was able to shift my comprehension from useless page counts to word counts. Though for some it might be like shifting from miles to kilometres.

The only problem with using a word count is that words vary in length, so two books with the same number of words can greatly vary the length of an e-book. One could have 50 screens, while the other is 60 screens.

[Solitaire1]

Originally Posted by Quoth

A trade size or mass market paperback "words" per page is better than "Letter Size", which is American anyway, for most ebooks and all novels. ISO similar size is A4.

The Adobe page numbers are independent of font, margins, screen size etc. Also ebook retail web sites often quote the number of pages, which is usually from the current paper edition.


Well, I get 95,000 words is about 280 pages. Some claim that words in English average 6 characters, but a word processor counts actual words. The differentiation of essay, short story, novella, novel, epic is usually by number of words, not characters. By 6 characters per word, my 280 pages for 95,000 words is maybe 570,000 characters.
Letter or A4 is too large.

Anyway, we don't need a new scheme.

I mentioned Letter (8.5 inches by 11 inches) only because it is a size that people in the U. S. are familiar with. Any paper size could be used in this calculation, such as A5 (literally half the size of A4).

[Karellen]

Originally Posted by Solitaire1

The only problem with using a word count is that words vary in length, so two books with the same number of words can greatly vary the length of an e-book. One could have 50 screens, while the other is 60 screens.

Sure, I suppose so. But I don't see the significance of that. If you know your book is 50,000 words, then what significance is "screens". What does it tell you that the word count has not already told you?

[JSWolf]

Originally Posted by madelonw

The Count Pages plugin doesn't count pages as they are numbered. In fact, the number it puts out is usually higher than actual print pages. I have my own columns for Fiction/Non-Fiction, and Category (i.e., novel, novella, short story, etc). I use the count pages result to say this is a novel, this is a novella, etc. It also has been useful in the past for selecting something to read that I will be able to finish during a 2-hour plane trip. Bottom line, the number is relative but I find it useful.

What page number type do you use with the plugin?

[JSWolf]

Originally Posted by Solitaire1

One possible way around this is to dispense with the conventional concept for page numbers all together and use an objective standard for dividing up an ebook that doesn't rely on variables such as screen size, size to the typeface, margins, and blank space around the screen. As an example, what if they defined a page in an ebook as 3,500 raw characters*. This is based on the number of characters (rounded down) on a sheet of letter-sized paper (54 lines at 6.5 inches wide each) using a fixed 10 pitch typeface. All that matters is the number of characters. Page 1 is from 0 to 3,500 characters, Page 2 is from 3,501 to 7,000 characters.

This, in turn would allow for a way to consistently calculate the number of pages in an e-book. By this definition, an ebook that has 1,000,000 raw characters would have a total of 286 pages (the last "page" would be a partial page).

*Raw characters means everything, letter, numbers, punctuation marks, formatting marks (such as tabs, line breaks, and hard returns) and spaces. The number of characters and spaces that can fit on a page (using what I described above) is exactly 3,510 raw characters.

We already have a page count method that doesn't rely on screen size, font used, line height, margins, font size, etc. That's the ADE page count method. Just use that for ePub.

[JSWolf]

Originally Posted by John F

I still like the way pbook pages were "counted": the story itself was what was counted, not the the other stuff (cover, copyright, acknowledgments, ...).

I do have those things counted. But things like HTML ToC, reviews, previews, publisher info, etc. I do delete so the page count I get is a lot more accurate because the rubbish won't get counted. Also changing things like <p class="indent"> to <p> and that helps make the page count more realistic.

[JSWolf]

Originally Posted by Solitaire1

The only problem with using a word count is that words vary in length, so two books with the same number of words can greatly vary the length of an e-book. One could have 50 screens, while the other is 60 screens.

Take the same eBook on the same Reader. One has a larger font size then the other. So one is going to have more page turns. I do sometimes up the font size when I am in bed and I want to finish the chapter but am getting tired.

But I do agree that number pf screens is not relevant when you are using ADE page numbers and/or word count.

[Quoth]

Originally Posted by JSWolf

We already have a page count method that doesn't rely on screen size, font used, line height, margins, font size, etc. That's the ADE page count method. Just use that for ePub.

Absolutely. They also have a page map method option to relate to a physical edition (needed certain text books). That can't be created automatically for an existing epub.