Report on new firmware?

[zelda_pinwheel]

Originally Posted by pilotbob

Ok, but just because it isn't there now, doesn't preclude it from happening. Perhaps some of the smarties here can get together and define a dictionary layout for the ePub format. Maybe FBReader or one of the open source readers will do it first.

BOb

actually nate the great has begun to do precisely that. i applaud him for it.

[Hajo]

Originally Posted by HarryT

No, it is an issue with the format, Bob. Mobi defines a dictionary format, and a mechanism for looking up words; ePub does not. There needs to be a specified dictionary format in order for dictionaries to be published in that format.

I can imagine a reader software that displays ePub documents and looks up selected words in a mobi dictionary - or in a database or in a ridiculously large txt file. No solution needs any changes to the current ePub format.

I really don't get, why an ePub reader needs a dictionary in an ePub format. As if the hyphenation information for MS Word has to be in DOC-Format, or the spell checker of an email client actually gets its word list out of an email. I assume that CorelDraw has spell checking. Where is this information stored? In a vector image?

[Godzil]

But for this, you will need support for DRM ePub AND DRM Mobi..

[HarryT]

Originally Posted by Hajo

I can imagine a reader software that displays ePub documents and looks up selected words in a mobi dictionary - or in a database or in a ridiculously large txt file. No solution needs any changes to the current ePub format.

I really don't get, why an ePub reader needs a dictionary in an ePub format. As if the hyphenation information for MS Word has to be in DOC-Format, or the spell checker of an email client actually gets its word list out of an email. I assume that CorelDraw has spell checking. Where is this information stored? In a vector image?

Well, you're right, of course, in that it doesn't NEED a dictionary in ePub format, but it needs a dictionary in some format that's available. The only eBook dictionary formats that I'm personally aware of are MobiPocket and Microsoft Reader (LIT) formats. We are all, unfortunately, well aware that Bookeen are not permitted to have both Mobi and ePub DRM formats in the same firmware, so that rules out Mobi dictionaries. That leaves LIT dictionaries, and no eInk device that I'm aware of implements LIT DRM, since that uses Windows cryptographics components which are not available on Linux (which the Gen3 runs).

So, what would your suggestions be?

[elena]

Originally Posted by Jellby

ePUB has embedded fonts,

why would this feature be a plus? It prevents me from choosing the font I prefer.

[pdurrant]

Originally Posted by elena

why would this feature be a plus? It prevents me from choosing the font I prefer.

That depends how it's implemented. And a bit on how far the format's been defined.

It should be possible to have a reader for ePUBs that allows you to over-ride any of the font choices in the ebook, and with a bit of luck, also identify the body copy font for you.

Embedded fonts are a plus for books with characters that aren't in most fonts. How about The Hobbit with an embedded runes font, or a maths books with a mathematical symbols font. Or even just a nice display font for chapter headings and/or drop caps.

[Snuffi]

Originally Posted by Hajo

I can imagine a reader software that displays ePub documents and looks up selected words in a mobi dictionary - or in a database or in a ridiculously large txt file. No solution needs any changes to the current ePub format.

I really don't get, why an ePub reader needs a dictionary in an ePub format. As if the hyphenation information for MS Word has to be in DOC-Format, or the spell checker of an email client actually gets its word list out of an email. I assume that CorelDraw has spell checking. Where is this information stored? In a vector image?

as others have already pointed out a mobi dictionary is not possible due to legal restrictions. A "ridiculously large txt file" ist not viable because it would be - well - ridiculously large. Performance would be a no-go for the road...

A database would certainly be possible but it would be pretty hard to implement DRM into a database and as Harry already pointed out no publisher would sell a dictionary without DRM. As much as I hope that DRM will die as fast as possible dictionaries will be certainly the last ebooks to be sold without copyright protection because there is a lot of work behind it and prices are high enough to make illegal copying appear interesting to many (I just assume that judging from what I see with music downloads: most people are willing to pay for music if they get it for a moderate price rather than search for pirated copies. But if they price is beyond a certain threshhold that willingness takes a drastic plunge).

So, although there are other possibilities, an epub dictionary would be the easiest to implement because the "infrastructure" is already there and it would be hard to convince the publishers to go for an all-new solution.

[Abelturd]

Originally Posted by HarryT

So, what would your suggestions be?

Someone has to start an initiative the goal of which will be to prepare the specs for epub dictionary format. Let's say IDPF in collaboration with european publishers and liseuse manufacturers (Bookeen, Endless Ideas, TXTR et al). Until the specification is finished and the dictionaries start being published to implement an open dictionary format like Stardict.

But the main idea is that manufacturers and publishers need to commence a mutual dialogue. The fragmentation of the market caused by the lack of standards, by the variety of reading devices supporting numerous ebook formats is doing noone any good.
So again -- COMMUNICATION!

[Godzil]

Abelturd: That's a really good idea

[Jellby]

Originally Posted by pdurrant

Embedded fonts are a plus for books with characters that aren't in most fonts. How about The Hobbit with an embedded runes font, or a maths books with a mathematical symbols font. Or even just a nice display font for chapter headings and/or drop caps.

Exactly. The latter is what I mainly use the feature for (see my Don Quijote upload, for instance). The main text has no particular font specified, so it will use whatever default the reader software has.

[HarryT]

Of course, if you just need a few equations, or a Greek quotation, or something like that, in a book, it can be done equally well (and more portably) with graphics - I've done that in many of the books that I've created. There are some books, however, that really cannot practically be created as eBooks without embedded fonts.

[Jellby]

Originally Posted by HarryT

Of course, if you just need a few equations, or a Greek quotation, or something like that, in a book, it can be done equally well (and more portably) with graphics - I've done that in many of the books that I've created.

I wouldn't say "equally well". Graphics don't wrap or scale nicely with font size. But for Greek (or Cyrillic, etc.), there are many fonts that include those characters, so they can be directly encoded (as unicode or entities) in the ebook and be displayed in whichever font the user has chosen. The Russian version of "La divina commedia" I posted was done that way, Greek would be just as easy, I believe.

[surfinjim]

Originally Posted by Abelturd

Someone has to start an initiative the goal of which will be to prepare the specs for epub dictionary format. Let's say IDPF in collaboration with european publishers and liseuse manufacturers (Bookeen, Endless Ideas, TXTR et al). Until the specification is finished and the dictionaries start being published to implement an open dictionary format like Stardict.

But the main idea is that manufacturers and publishers need to commence a mutual dialogue. The fragmentation of the market caused by the lack of standards, by the variety of reading devices supporting numerous ebook formats is doing noone any good.
So again -- COMMUNICATION!

Dear Abelturd,

As the technical architect of a company which is developing embedded software for ebook devices, I'm following your discussions with pleasure and interest. I hope we will soon have time to add dictionary support in our Epub reader. So please don't hesitate posting your suggestions here, they'll be studied ;-)

Best regards,

Surfin'Jim

[HarryT]

Devising a standard is the easy part. Persuading Oxford, Collins, Chambers, Mirriam-Webster and all the other dictionary publishers to publish their dictionaries in that format, as they currently do in Mobipocket format; THAT is the challenge, and will, I fear, take years before it happens. And that's once you manage to get a dictionary format standard adopted by the ePub standards committee, which itself is going to be a far from trivial undertaking.

[junior]

Originally Posted by HarryT

No, it is an issue with the format, Bob. Mobi defines a dictionary format, and a mechanism for looking up words; ePub does not. There needs to be a specified dictionary format in order for dictionaries to be published in that format.

As there is no mobi dictionary English => Portuguese.

I made my own dictionary.

Mobi dictionary have a specific format.

The mobi format have a own engine search while EPUB not have a engine search.