Expansion of Book Uploads for the iLiad & for other devices

[JSWolf]

Since the iLiad had Mobipocket for it, just make the conversions for it in PDB format and then the palm types can read it as well. LIT format is good to have for use on the computer or laptop. And yes there are good tools to convert from LIT to something else such as LRF. Book Designer would be the best tool of choice in this multi-conversion as it will allow format once, write multiple.

Jon

[Robert Marquard]

I meant the HTML download as an additional option. The HTML is the base for most ebook formats so it is needed for those who want to create the various formats. I doubt that all the creators here know all the tricks for all the formats equally well. So collaboration is needed. Probably the download area needs to be moved over to the wiki.

[JSWolf]

Originally Posted by Robert Marquard

I meant the HTML download as an additional option. The HTML is the base for most ebook formats so it is needed for those who want to create the various formats. I doubt that all the creators here know all the tricks for all the formats equally well. So collaboration is needed. Probably the download area needs to be moved over to the wiki.

The solution to that would be easy. When you attach the book for the forum, make it a ZIP file with the book already compiled and the source file be it HTML, HTML0, or whatever was used. Include a text file to say what program was used and the options and there you go.

[rlauzon]

Originally Posted by NatCh

Perhaps it'd be better to upload the books to FeedBooks for processing into Custom PDFs?

FeedBooks seems to only offer eBooks in PDF form (according to the quick check I did), making the service worthless.

I can understand offering eBooks in ready-to-go formats for those who don't want to be bothered to do their own conversion, but the original source should [B]always[B] be offered as well.

[Hadrien]

Originally Posted by rlauzon

FeedBooks seems to only offer eBooks in PDF form (according to the quick check I did), making the service worthless.

I can understand offering eBooks in ready-to-go formats for those who don't want to be bothered to do their own conversion, but the original source should [B]always[B] be offered as well.

I don't think that all of our registered users consider Feedbooks worthless, hopefully. PDF is what we selected for the beta, since it's the only good way to keep nice typesetting in an e-book for the moment.
What you call original source is actually quite different on Feedbooks: the source is not a file but in multiple tables in a database, divided into elements that are not presentation tags but parts/chapters/sections.
And some of the presentation tags are book specific too like footnotes for example.
Oh and unlike many websites, the source is always included both on the website and in the book, so you can easily find who scanned, did the proof reading etc...

Based on your opinion, I guess that you also consider the LRF uploads worthless, since on the forum, people usually upload the LRF only ?

[rlauzon]

Originally Posted by Hadrien

What you call original source is actually quite different on Feedbooks: the source is not a file but in multiple tables in a database, divided into elements that are not presentation tags but parts/chapters/sections.

Then you should be able to provide eBooks in a non-restricted format just as easily as you can in PDF.

Originally Posted by Hadrien

Based on your opinion, I guess that you also consider the LRF uploads worthless, since on the forum, people usually upload the LRF only?

All closed, proprietary formats are worthless.

[yvanleterrible]

We should have the base books cleaned prepared and stored in HTML. When you click on a book choice a dialog box asks you what format you need and a universal translator coughs it up for your download. That would be a neat service similar to a vending machine and necessitating less server space. MR's service would stand out with the offer of tasteful displays like those RWood and HarryT do. Every format available could be provided, particularly orphaned device formats, those that are no longer supported by their departed makers.

Can this kind of software be done and offered on a large scale?

[Hadrien]

Originally Posted by rlauzon

Then you should be able to provide eBooks in a non-restricted format just as easily as you can in PDF.

We can and we will be supporting Open Ebook in the future. But for us, HTML/XML is not the source.

Why ?
Because if we only used HTML/XML as the source, we wouldn't have any markups dividing the book correctly. A book is not strictly presentation, there's also semantic information in it. HTML is not the nexus for e-book: it lacks some very valuable information.
We must not represent a chapter as something slightly bolder and on the center of the page, in what would be the source of all those e-books formats. You don't need to store a TOC in the source, you can easily create a TOC if you have the right markup for parts/chapters/sections.
The same thing could be said for footnotes too.

Restrictions are not purely based on DRM or reflowable capabilities.

[NatCh]

Originally Posted by rlauzon

FeedBooks seems to only offer eBooks in PDF form (according to the quick check I did), making the service worthless.

I can understand offering eBooks in ready-to-go formats for those who don't want to be bothered to do their own conversion, but the original source should [B]always[B] be offered as well.

(Sigh) I was talking specifically about how we might handle uploads of PDF books, rlauzon, so the fact that FeedBooks only offers the thing I was specifically talking about is kinda the point.

I know you don't like PDF, Mobi, LRF, and all the other 'closed, proprietary' formats, and I'm not disputing the point, but a lot of folks do get a lot of mileage out of them.

I think that when we're talking about ways to offer free books in as many formats as possible, for the convenience of as many folks as possible, with as little trouble as possible, the fact that some of those formats are terminal ones (as in you can't get the stuff back out) is not really the most important point in that context.

[rlauzon]

Originally Posted by Hadrien

Because if we only used HTML/XML as the source, we wouldn't have any markups dividing the book correctly.

This is untrue.

You are correct that HTML may not provide all the information you need, but you are incorrect when you say that XML cannot. XML is often used as the "flattened" form of a database.

Originally Posted by Hadrien

A book is not strictly presentation, there's also semantic information in it.

Which XML can provide.

[rlauzon]

Originally Posted by NatCh

(Sigh) I was talking specifically about how we might handle uploads of PDF books, rlauzon, so the fact that FeedBooks only offers the thing I was specifically talking about is kinda the point.

And when the next eBook reader comes out that doesn't support any of the PDF eBook sizes, all the work you did becomes useless.

Originally Posted by NatCh

I know you don't like PDF, Mobi, LRF, and all the other 'closed, proprietary' formats, and I'm not disputing the point, but a lot of folks do get a lot of mileage out of them.

I do understand. But I also understand that as PDF (or any other locked down format) gains users, it makes it that much harder for a standard eBook format to be created.

[NatCh]

Originally Posted by rlauzon

And when the next eBook reader comes out that doesn't support any of the PDF eBook sizes, all the work you did becomes useless.

Which is why I suggested handling them through FeedBooks which creates them on the fly and can re-create them for other sizes, eliminating the human effort from the equation altogether.

Originally Posted by rlauzon

But I also understand that as PDF (or any other locked down format) gains users, it makes it that much harder for a standard eBook format to be created.

I'm not disputing that either, but we weren't discussing what formats should be used in a universal sense, but rather how to best serve our members, some of whom use terminal formats such as PDF, and I see FeedBooks as an excellent way to not put a lot of human effort into creating the individual books. Hadrien has already put a lot of his own effort (which I assume is human ) into creating an engine that does the work for us.

Yes, he's working on handling other outputs, but it's not a finished product yet.
Yes, there are other ways to do it than the way he chose, but he did have to choose one.
Yes, we all want a single, standard format, but we don't have one yet, and people have to choose from the ones available, or give up e-reading altogether, which doesn't help further e-books either.

That's all I'm trying to get across here.

[Hadrien]

Originally Posted by rlauzon

This is untrue.

You are correct that HTML may not provide all the information you need, but you are incorrect when you say that XML cannot. XML is often used as the "flattened" form of a database.



Which XML can provide.

Yes and we're going to use XML for this. But most XML-based formats are purely based on presentation and nothing else (that's what I was pointing to, using XML in my previous post). In the future we'll use as an input xml that will be divided into 2 different types of tags:
- metadata and structure
- presentation

Metadata will be for the name of the book, the title, the isbn, the tags, short description etc...
Structure will divide the the book into parts/chapters/sections and might be extended for other type of texts (plays ?).
Presentation will keep some basic XHTML tags (b, i, br, center, hr...) but we'll also add a few book specific tags (<fb:footnotes>, <fb:verse> for example).

With such an XML file we'll have exactly the same stuff than in our database.

Out of this structure, we'll create files on the fly like we already do (PDF right now, XML based formats in the future, we already tested TTS too).

In our admin, we already have a feature to export our book in XML form, but very basic support right now, we'll work on extending this during the summer.

[yvanleterrible]

@ rlauzon

What's your idea of the best format we should have? Is there one now you'd see over others?

[NatCh]

That sounds like a good topic for it's own thread, yvanleterrible.