Do you consider PDF to be a legitimate "e-book" format? Why or why not?

[Liviu_5]

Originally Posted by kovidgoyal

Scientific publications certainly haven't moved from latex to pdf, they're still latex, PDF is simply the latex rendered. And eventually, hopefully, MathML will mean that even scientific documents can be rendered in a reflowable manner.

You are absolutely right from a technical point of view (many people still write in Latex since it's such a cult editor, and of course it's free - when I was in academia I hated it ), but I remember not that many years ago when I had to install a dvi viewer and maybe even a latex compiler to read the papers of interest to me (I still like to keep up with this or that development even though I left academia many years ago) while today pretty much anything is available in pdf, so in a measurable way pdf made lots of inroads.

Now I do not particularly like pdf and positively loathe it for text based fiction where it's not needed, but I think it's here to stay for a while whatever its weaknesses.

[akiburis]

That's an interesting point of view, akiburis. Would you mind elaborating on where you see those other formats as being lacking compared to PDF? I'm less sensitive to such nuances myself (I figure if I can read the words clearly without moving my lips, I'm doing pretty well ), and I'd like the perspective of someone who is more aware of them.

Well, for example, their inability to render properly spaced and hyphenated justified text (or, for that matter, properly spaced and hyphenated unjustified text). This really does make a difference for me in the pleasure and ease of reading. I dislike reading ragged-right text generally and wildly ragged text always, and justified text without hyphenation is pretty unpleasant too---at least, I'm not inclined to settle for such things if I can have something better. I don't think this is a matter of being especially sensitive to nuances.

My perspective may differ from many people's in that I have no particular preference for ebooks per se versus paper books. But there's lots of digital text around, LCDs do not make for comfortable sustained reading, and printing the stuff out is expensive and unsatisfactory (my cheap laser printer isn't a printing press and bookbinding machine, after all); so I find a device like the Sony reader really useful for displaying digital text. Often, if it's worth it, formatted (typeset) with LaTeX and PDF output. Often the plain text is enough for my purposes or the plain text minimally formatted in Word and converted to RTF. But rather than depend on the Reader's rather weird plain text and RTF handling, I export that stuff to PDF, too, before loading it onto the Reader, to preserve the formatting. That's exactly what PDF is useful to me for, and why the device would be more or less useless to me without decent PDF support.

Now, what gets my goat about all this anti-PDF rant is that it seems to be meant to deprive me---if the ranters (excuse me) had their way---of something that I find very useful and, at present anyway, irreplaceable. And the anti-PDF arguments are all hooey. PDF is, everyone incessantly insists, a terrible format for ebooks because it was never meant for ebooks and is a terrible format for ebooks---a circular, question-begging argument if ever there was one. Or, because ebooks are a new digital technology, they must by some technological imperative be different in every way from paper books---that is really what a lot of these arguments come down to, it seems to me. But a technogical advance or change in the means of production and distribution---which is what the invention of movable type was, for example, and what digital type may be---needn't entail any essential change in the form of the article produced. In either case, however, it's assumed, not in any way demonstrated, that an ebook must be such-and-such, and it follows that no one must want or be allowed to have any other sort of ebook or ebook format or ebook device. All petty totalitarian fantasy, in my view.

[tompe]

Originally Posted by akiburis

This really does make a difference for me in the pleasure and ease of reading. I dislike reading ragged-right text generally and wildly ragged text always, and justified text without hyphenation is pretty unpleasant too---at least, I'm not inclined to settle for such things if I can have something better. I don't think this is a matter of being especially sensitive to nuances.

It should be possible to implement TeX's hyphenation algorithm in a readers program. Strange that nobody has done it.

Justified text without hyphenation on the Gen3 works rather well since the reader program accepts that some lines are not justifed if justfication would lead to too bad spacing. And that is actually not so disturbing.

Ragged text I have not tested since as you point out ragged text without hyphenation is worse than justified text without hyphenation.

[kovidgoyal]

Originally Posted by tompe

It should be possible to implement TeX's hyphenation algorithm in a readers program. Strange that nobody has done it.

I have in my LRF viewer.

[Liviu_5]

Originally Posted by tompe

It should be possible to implement TeX's hyphenation algorithm in a readers program. Strange that nobody has done it.

Actually Fbreader does hyphenation (have no idea what algorithm though) in several languages at that.

[NatCh]

Originally Posted by akiburis

Well, for example, their inability to render properly spaced and hyphenated justified text (or, for that matter, properly spaced and hyphenated unjustified text). This really does make a difference for me in the pleasure and ease of reading. I dislike reading ragged-right text generally and wildly ragged text always, and justified text without hyphenation is pretty unpleasant too---at least, I'm not inclined to settle for such things if I can have something better. I don't think this is a matter of being especially sensitive to nuances.

Perhaps I expressed myself poorly, I did not mean to suggest that you were oversensitive, only that I was less sensitive to such things. Perhaps "inattentive" would be a better description of my own perceptions.

In any case, I see now what you're saying about the greater care for the actual layout that you get with a well-done PDF, and I'd agree with you that it's cleaner than what you get form most (if not all) e-book reading apps. It's a valid point, even though I find it less compelling than you do, if only because that careful work would need to be done for each target display size. It may not be a big deal for one size, but it gets pretty burdensome for multiple sizes.

Originally Posted by akiburis

My perspective may differ from many people's in that I have no particular preference for ebooks per se versus paper books. But there's lots of digital text around, LCDs do not make for comfortable sustained reading, and printing the stuff out is expensive and unsatisfactory (my cheap laser printer isn't a printing press and bookbinding machine, after all); so I find a device like the Sony reader really useful for displaying digital text. Often, if it's worth it, formatted (typeset) with LaTeX and PDF output. Often the plain text is enough for my purposes or the plain text minimally formatted in Word and converted to RTF. But rather than depend on the Reader's rather weird plain text and RTF handling, I export that stuff to PDF, too, before loading it onto the Reader, to preserve the formatting. That's exactly what PDF is useful to me for, and why the device would be more or less useless to me without decent PDF support.

Yes, it's extremely useful as you describe. There are trade-offs to that approach (just like everything else in the world), but you've clearly decided that what you lose is less important to you than what you've gained. Good for you.

Originally Posted by akiburis

Now, what gets my goat about all this anti-PDF rant is that it seems to be meant to deprive me---if the ranters (excuse me) had their way---of something that I find very useful and, at present anyway, irreplaceable.

Great googely-moogely, no! Not in the least do we want to abolish PDF format. Shucks, I don't even want to remove support for it from the Reader, or anything else! It is, as you say, irreplaceable for certain purposes. I don't think anyone is arguing for its abolition.

This wasn't even meant to be a rant! Only a discussion of the suitability of PDF format as an e-book format. I phrased it like I did to catch interest and get folks discussing the matter to see what interesting perspectives might come out of it, that's all.

Originally Posted by akiburis

And the anti-PDF arguments are all hooey.

Well, that's certainly a matter of opinion. I did mention that the whole discussion was within the realm of opinion, didn't I?

Certainly, you have every right to consider the views of others uncompelling, but they have as much right to hold them as you do to not hold them, right?

Originally Posted by akiburis

PDF is, everyone incessantly insists, a terrible format for ebooks because it was never meant for ebooks and is a terrible format for ebooks---a circular, question-begging argument if ever there was one. Or, because ebooks are a new digital technology, they must by some technological imperative be different in every way from paper books---that is really what a lot of these arguments come down to, it seems to me.

I'm sorry, but you lost me here, I've not seen anyone arguing any of this.

Seems to me the majority "no" opinion that has been expressed is more that the undeniable strengths that PDF has, the things it does so well, aren't compatible with what those particular folks want from an e-book. That is to say that the excellent job it does of preserving formatting is less important to those individuals than the ability of an e-book to adapt to whatever display size is being used.

I'm pretty sure I haven't seen anyone suggesting that PDF is too old-fashioned to be used for e-books.

Originally Posted by akiburis

In either case, however, it's assumed, not in any way demonstrated, that an ebook must be such-and-such, and it follows that no one must want or be allowed to have any other sort of ebook or ebook format or ebook device. All petty totalitarian fantasy, in my view.

Hardly! This entire discussion is, at its core, a facet of what people want their e-books to do, not what they must be. Yes people are speaking from pre-conceptions, but I don't think those preconceptions are assumptions of what e-books should or must be. In any case, I don't see where anybody is suggesting that there can only be one solution.



I'm very sorry if we gave that impression, I really don't believe it was intended. I can only speak definitively for myself, of course, but I most definitely did not mean to suggest any such thing. I apologize for inadvertently giving that impression, as I evidently have.



In any case, please, let's all now consider it expressly appended to the whole premise of the discussion that we're not aiming to abolish nor institute any given format as a single solution within the confines of this discussion.

If anyone does want to have that discussion, start another thread, there's plenty of space for another one.

[JohnClif]

PDF is not a valid ebook format, although it is most likely the most predominant ebook format.

The problem with PDF is that it is a rendering format, showing the content at one fixed page size. If the ereader display is smaller than the fixed page size, either the document is scaled or a viewport onto the document must be displayed... and neither option is optimal. Nearly all PDF ebooks are formatted for a page size that doesn't conveniently display on typical devices such as mobile phones or dedicated ereaders like the Kindle or Sony. And, because PDF is inherently an image format (scalable vectors), it is hard to convert to text while saving document formatting, as witnessed by the bad job most PDF-to-HTML converters do.

Hypertext, in all its variants, e.g., HTML, .MOBI, Kindle's .AZW, etc., is a display format, allowing the content to flow to fit variable page sizes or changing font sizes. A full 'page' depends on a combination of the font size and display size. Because hypertext is based upon textual data (instead of scalable vectors), it is easy to adjust the layout for difference screen and font sizes.

I'm sure someone is working on the perfect PDF-to-hypertext converter... and it will make them a lot of money.

[Hadrien]

Originally Posted by Liviu_5

Actually Fbreader does hyphenation (have no idea what algorithm though) in several languages at that.

It's TeX's algorithm.

[JohnnyD]

Not considering drm, I think pdf is a fine e-book format.

There is a bit of an analogy between pdf and html: if you want your webpage to be viewable on every webbrowser, keep your content in clean html. But if you want a really spiffy looking webpage, you're going to need a lot of bells and whistles, which are definitely not html... meaning your webpage will only be viewable on some fairly recent webbrowsers...

I guess the same goes for e-books. If you want your story to be read by everybody on every e-book reader, make it available for free in an open format (html?). But if you have some really specific layout and design needs, it might be that pdf is the only way to go...

So, is pdf a valid e-book format? It depends, but I guess it has its place...

[yvanleterrible]

For PDF to work on a portable device only means you need basic software and enough computing power to work properly. It is not yet part of most maker's incentives to allow now. They all have their own formats and agendas to fulfill. At least Sony shows it's willing.

[ath]

It is probably the only e-book format that allows us to implement what the past 400 years of typesetting have taught us about what makes a page legible as well as readable: proper line length, proper selection of typeface, proper presentation of material on the page, and so forth. Reading should be unconscious to be considered good. And that takes a lot of work and knowhow to achieve and maintain.

Complaints that PDF is locked to a fixed page format can safely be ignored: p-books are also locked to a fixed format, but there are no serious complaints about those. I choose to regard this particular objection more as an indication that current e-readers (regardless of publishing format) are the wrong size, or bad in some other important way, such as poor resolution or contrast (more probably both), that need to be mitigated. It does not make sense to find a best format for bad e-readers: the only thing that makes sense is to evaluate the badness.

And complaints that someone only wants to have a particular typeface to read can also be safely ignored. Typically it's another manifestation of poor page size/resolution/contrast, and I've already covered that. It sometimes happens that it's taken as a sine-qua-non of user configuration. User preferences are fine to support, but they should not be allowed to take over the shop. It's better to develop some reading musculature instead.

When we have a reader device that does the page size of a King Penguin volume, and with at least the same contrast and resolution of one, I'd say this field has ... not matured, but reached an acceptable lowest common denominator. Until then, all we're doing is stop-gapping. And there are a lot of possible ways to do that -- as long as we don't begin to think that that particular activity is especially important, it doesn't matter.

Note: I'm talking about reading books here, and the functionality of an *e-book* format (i.e. I take the term very literally indeed). Not newspapers, not the latest weekly/monthly/yearly report, or the last report printouts to proofread for the next version. Just books.

[Penforhire]

A lot of what you are arguing for does not have to be included in the actual e-book format. Why can't the rendering device perform proper kerning, line length, hyphenation, typeface choice, etcetera?

[nekokami]

A fair amount of the non-fiction I read (including books, not just journal articles) has math in it. While there are non-PDF (and non-TeX) solutions for rendering mathematical formulae, I think we're going to continue to see TeX->PDF for a while yet.

Edit: It's also handy for scanned documents, including books.

[NatCh]

Originally Posted by Penforhire

A lot of what you are arguing for does not have to be included in the actual e-book format. Why can't the rendering device perform proper kerning, line length, hyphenation, typeface choice, etcetera?

I don't see any reason that it couldn't -- aren't there several small word-processing apps that do so? It seems to me like that particular wheel has been thoroughly invented

[kovidgoyal]

@ath: You seem to be coming from the perspective that the job of an ebook is to be a pbook. If that's you're perspective, then you're going to love PDF. If however, like me, you believe the job of the ebook is to make the pbook obsolete, then PDF is simply not the answer.