EPUB3 is dead, long live EPUB3Lite...

[Katsunami]

Why exactly do we need an all-singing, all-dancing EPUB version? IMHO, EPUB as it is now is good enough. If more is needed, implement/support more CSS.

E-books don't need video's, or sound, or multimedia. Use other formats for that. Call it MPUB for all I care. Use EPUB for the e-books, and MPUB for multimedia stuff.

[rkomar]

I, too, would like to see support for MathML in epub. It is more-or-less required for technical documents. Including tiny images of equations is a poor replacement.

Originally Posted by Katsunami

Why exactly do we need an all-singing, all-dancing EPUB version? IMHO, EPUB as it is now is good enough. If more is needed, implement/support more CSS.

I think a lot of it comes down to the just-one-more-feature mentality. For example: MathML would be handy for technical books. I've seen mention of sheet music as well in these forums. I'm pretty sure the desire for multimedia elements comes from textbook publishers, who could use it as a selling point for their product (and it is sometimes a justifiable selling point).

Now if you're looking at ebooks and saying that they're just for novels, or certain types of non-fiction, then yes the current specification is probably fine. If you're looking at ebooks with the desire to expand their potential (even if it's just to the point of print books), then a lot of work has yet to be done.

[ChrisJakarta]

For those of us who just want to publish documents and booklets for free-of-charge distribution and access on a wide variety of devices, what is the best option to provide reasonable formatting and possibility of adding graphics? I just loved the old Palm Markup Language, but I guess that's no longer an option. I've tried ePub, but none of the reader programs seem to fully implement it, so the result is often a dog's breakfast...

Chris

[tubemonkey]

Originally Posted by fjtorres

But I'm not holding my breath for an outbreak of sanity.

LOL

[Rob Lister]

Originally Posted by rkomar

I, too, would like to see support for MathML in epub. It is more-or-less required for technical documents. Including tiny images of equations is a poor replacement.

What about a non-tiny image? An image of a formula can be as large as the screen of the reader. It can be rendered in either absolute or relative terms. It can be as crisp and concise as any font.

I can see that it would be useful to include the creation functionality in epub production software, but in terms of rendering, I don't see why a jpeg would not suffice.

I'm not all that familiar with MathML (I always meant to learn it but I was dropped on my head as a baby) so explain what I'm missing.

[HarryT]

Originally Posted by Rob Lister

I can see that it would be useful to include the creation functionality in epub production software, but in terms of rendering, I don't see why a jpeg would not suffice.

You wouldn't use a JPEG, you'd use an SVG image so it scaled with the text.

[fjtorres]

What the academic world wants is an evolved version of Mathcad where equations in a document aren't images or even characters/symbols but live equations that can be manipulated in-situ to do symbolic math or numeric calculations

http://en.wikipedia.org/wiki/Mathcad

The sort of thing you can do with Mathematica or embedded mathcad objects in an MS WORD document.

Until that can be delivered, they'll be Oliver Twist, always pleading for "more?"

[HarryT]

Originally Posted by fjtorres

What the academic world wants is an evolved version of Mathcad where equations in a document aren't images or even characters/symbols but live equations that can be manipulated in-situ to do symbolic math or numeric calculations

http://en.wikipedia.org/wiki/Mathcad

That's a totally different thing to MathML, which is what we're talking about here.

[fjtorres]

Originally Posted by HarryT

That's a totally different thing to MathML, which is what we're talking about here.

I know.
But that is what they really want.
MathML is merely a way to get them closer to it.
But even with MathML they won't be satisfied.
What is really needed is a wholly separate format for academia instead of trying to bloat up a trade publishing spec.

"One ring to rule them all" never ends well for anybody.

[murraypaul]

Originally Posted by Rob Lister

What about a non-tiny image? An image of a formula can be as large as the screen of the reader. It can be rendered in either absolute or relative terms. It can be as crisp and concise as any font.

I can see that it would be useful to include the creation functionality in epub production software, but in terms of rendering, I don't see why a jpeg would not suffice.

Because the equations should properly flow to fit the space available, not just be a statically rendered image scaled to different sizes.

You would ask this question, would you?:
"Why do you need reflowable text formats? Why not just take a PDF file and zoom in as you need to?"

[graycyn]

Originally Posted by HarryT

You wouldn't use a JPEG, you'd use an SVG image so it scaled with the text.

How do you actually make an SVG image?

I have this book I'm working on and there is this text cipher and the only way I could figure out to work it is either in a table (and worry about it breaking in some app) or as an image, so I made it a PNG. But of course that will not scale with the text.

I'm clueless about SVG images and how they work or how you put them in the HTML, etc... all I know about is the SVG cover wrapper, but I don't need the cipher to be full page.

[WillAdams]

Equations should also be allowed to break --- I haven't found a tool yet which will allow me to automatically / efficiently indicate break points in an equation when making a .svg.

Originally Posted by fjtorres

MathML is merely a way to get them closer to it.
But even with MathML they won't be satisfied.
What is really needed is a wholly separate format for academia instead of trying to bloat up a trade publishing spec.

I'm not so certain about that. There are a number of commercial and open source formats that already do that.

I would think that there are two goals to MathML: one is to create device independent markup for formal languages. The second is this whole idea of a semantic web, or encoding things in a way that reflects what it means. Software may be able to manipulate that encoded information in the ways that you're suggesting, but that is up to the tools accessing that information, and not a part of the specification.

[Freeshadow]

They should have taken a look at how TeX and co. is done IMO.