Mobipocket vs ePub: Why worse is better in ebook formats

[Ken Maltby]

This may be unfair in some regard, but here is a simple comparison:

On the Ectaco JetBook Lite (JBL) with the 0.16e FW and the Adobe Mobile Reader
for epub and (I think) a version of the FBReader for .azw/.mobi, a comparison of the same
ebook ("The New Road To Serfdom"). With the book disinfected after purchase and the
resulting .azw renamed to .mobi and converted to .epub in Calibre. Both files added to
my master SD card directory, in the "PolyTics" folder on my PC hard drive and synced to
my SD Card with SyncToy.

The .mobi file is at 407KB while the .epub is only 238KB. The JBL reading of the .epub
includes the cover page the .mobi does not. The page formatting and general
appearance of the .epub closely matches the display of the book in the PC Kindle
Reader, the .mobi looks very primitive in comparision and lacks a lot of the formatting
present in the PC Kindle Reader's display of the ebook.

Now the point of this is that the JBL is not a high powered device (except in comparison
to the Ectaco Mini, perhaps.) There is a great deal more given up by a simplified readers
implementation of the .azw/.mobi than is given up by the .epub implementation.
Remember the much better looking .epub was created from the same file as the .mobi.
(The .mobi is just the .azw file renamed.) All the same layout and formatting data is there
in the .mobi file, but not implemented.

As an example of a practical experience, at least with the current JBL firmware, the ePub
is outshining the .mobi by a large margin.

Luck;
Ken

[karunaji]

Wasn't one of the epub deficiencies a lack of dictionary format? Are there any progress in this regard?

[murraypaul]

Originally Posted by karunaji

Wasn't one of the epub deficiencies a lack of dictionary format? Are there any progress in this regard?

The format the dictionary is stored in is surely irrelevant?
ePub readers can certainly include dictionary lookup.

[karunaji]

Originally Posted by murraypaul

The format the dictionary is stored in is surely irrelevant?
ePub readers can certainly include dictionary lookup.

The Kobo reader doesn't provide such option yet. Kindle does this out of box, probably, because it is defined by the mobipocket format. And you can use the same dictionaries that are already published, if we don't count DRM issues.

If there is an epub reader which allows to use 3rd party dictionaries, then my existing mobi dictionaries need to be converted to whatever format this device can use. I guess that single defined format would provide some incentive to epub reader developers to implement dictionary support sooner than later.

[murraypaul]

Originally Posted by karunaji

I guess that single defined format would provide some incentive to epub reader developers to implement dictionary support sooner than later.

iBooks already supports dictionaries, don't know about other readers.
It makes sense that you would store a dictionary in a specialist format designed for quick lookup, rather than a generic ebook format. A dictionary is a database, not a book.

[karunaji]

I would say that it is a special type of book that is also a database. As dictionary entries contain the same elements as any other book (different fonts, text formatting, pictures, audio, links to other entries etc.) it becomes an overlay to any ebook format. Some elements are even global, like style sheets. I hope that epub format designers have anticipated such use.

[dmikov]

Originally Posted by murraypaul

iBooks already supports dictionaries, don't know about other readers.
It makes sense that you would store a dictionary in a specialist format designed for quick lookup, rather than a generic ebook format. A dictionary is a database, not a book.

So is a Nook and new Sony Reader so it's not book format question but reader.

[DaleDe]

Originally Posted by karunaji

I would say that it is a special type of book that is also a database. As dictionary entries contain the same elements as any other book (different fonts, text formatting, pictures, audio, links to other entries etc.) it becomes an overlay to any ebook format. Some elements are even global, like style sheets. I hope that epub format designers have anticipated such use.

When displayed it looks like a book but it is not a book when used as a look up dictionary. There has to be a rapid indexing mechanism that leads directly to the area of interest and it needs to be displayed as an overlay to whatever book you are reading, thus it looks like a book but does not act like a book. It is a database that displays like a book.

Cool Reader has dictionary support and can also read ePUB files so any eBook Reader that support Cool Reader also can support dictionaries so long as you use Cool reader to read the book. However, a full implementation of dictionaries on an eBook reader should be able to use the dictionary no matter what the format of the book is in. This is why ePUB format should not be a requirement for support of dictionaries. It needs way more specific support requirements that current ePUB in order to be an exchange format for dictionaries.

Dale

[thrawn_aj]

Originally Posted by menneske

just as long as it's not poetry. Breaking the lines and making prose of meticously designed poetry basically kills poetry.

Agreed. In fact, I asked my college English teacher (after a particularly non-rhyming poetry session ) what defines poetry these days. Her answer cracked me up - "Poetry is solely defined as writing that contains hard line breaks".

[DaleDe]

Originally Posted by thrawn_aj

Agreed. In fact, I asked my college English teacher (after a particularly non-rhyming poetry session ) what defines poetry these days. Her answer cracked me up - "Poetry is solely defined as writing that contains hard line breaks".

Poetry can be problematic when the lines are as short as they are in an eBook Reader. I think the important thing is to expect that the lines may be folded and thus the best thing is to designed paragraphs that have no space between them and then Outdent the first line of the paragraph. This way the first line will be furthest to the left. Each line begins a new paragraph. All lines will line up and any overflow due to long lines will appear indented. Stanzas will have space between the lines of the verse.

To me this represents the best guideline but of course some styles use indenting and spacing as part of the poem and this may require special handling to keep the poem intact.

Dale

[karunaji]

Originally Posted by DaleDe

When displayed it looks like a book but it is not a book when used as a look up dictionary. There has to be a rapid indexing mechanism that leads directly to the area of interest and it needs to be displayed as an overlay to whatever book you are reading, thus it looks like a book but does not act like a book. It is a database that displays like a book.

If it quacks like a duck...

Dictionary is used as a reference but much more than just a quick overlay. There should be a standard way to quickly link to all dictionary entries so that reader software in question can index the dictionary. The problem with iBooks is that their dictionary format is unknown, thus there is no way I can make my own dictionary. It has no dictionary support, it simply has some in-built dictionaries.

Even if they supported 3rd party dictionaries in some proprietary format, I wouldn't want to buy or create a new dictionary for each device just because the developer used a slightly different scheme. Probably all it needs is defining some tags based on epub format and calling it epub dictionary specification or extension.

[menneske]

Originally Posted by DaleDe

Poetry can be problematic when the lines are as short as they are in an eBook Reader. I think the important thing is to expect that the lines may be folded and thus the best thing is to designed paragraphs that have no space between them and then Outdent the first line of the paragraph. This way the first line will be furthest to the left. Each line begins a new paragraph. All lines will line up and any overflow due to long lines will appear indented. Stanzas will have space between the lines of the verse.

To me this represents the best guideline but of course some styles use indenting and spacing as part of the poem and this may require special handling to keep the poem intact.

Dale

The programmer of Calibre has programmed somthing in html, which can be implemented in handling of Poetry (can't find the link just now, but it is an epub, that deals with Dropcaps, Floats, Full screen image and Poetry - should be possible to find)

It works by indenting the words that are broken off at the right margin, but l don't know what happens, if a line is already indented. I myself (as a poet) use indentions/tabulations as a defined pause, longer than comma and with a build-in "holding of breath".

I don't yet know how to adjust EPUBs, so I can't test it.

anyone?

ex.

I made this line appear
out of nowhere
out of nowhere it came
so sudden
so sodden
suddenly
from a simple word I entered into it in order to appear wise and experienced (here the line breaks in this window)
This can be changed to beauty
beautiful now it is
beautiful
and calm

[Ben Thornton]

I read somewhere (I think that it was a LeGuin essay) the view that prose was about communicating content, whereas for poetry, form was equally important.

Hence the importance of formatting, I suppose.

[Shaggy]

Originally Posted by murraypaul

The format the dictionary is stored in is surely irrelevant?
ePub readers can certainly include dictionary lookup.

They may be able to, but the main ones don't. The Adobe based one (which is what many devices use), for example, doesn't have it. Whether that's just because they chose not to, or there is something in ePub itself that makes it difficult, I don't know.

[DaleDe]

Originally Posted by Shaggy

They may be able to, but the main ones don't. The Adobe based one (which is what many devices use), for example, doesn't have it. Whether that's just because they chose not to, or there is something in ePub itself that makes it difficult, I don't know.

Dictionary lookup is too slow to be useful if it is just a document. you need an index of words to look up before it is useable.


Dale