Project Gutenberg going Mobi?

[mtravellerh]

Today I noticed something peculiar: There is a new format supported by Project Gutenberg. Besides Epub (we have discussed this in other thread) there are now some books in Mobi format. Up till now I haven't seen any official news from PG concerning this astonishing move. Mobis (illustrated and non illustrated) can be found, among others, in "Why I Believe in Scouting for Girls by Mary Roberts Rinehart" and in "Ben, the Luggage Boy; by Horatio Alger", both books featuring an illustrated and non-illustrated Mobi version.

On first view, the books seem quite nicely formatted, with the usual blurb before and after the body of the book. Covers, however, are correctly defined and display outside of the book body (they appear in the book body a second time).

No such luck for TOCs. The inline TOCs seem to be working (so far) but no outside one (it would have been quite easy via a simple anchor definition, though).

The books are justified and do have correct metadata as far as I could see. We don't know anything about the scale of the Mobi integration yet, but we hope that this evolution goes parallel with the Epub integration. There is for sure something moving inside PG.

Update:

In the format explanation, PG gives this short statement:

Important notice: Mobipocket files are a new experimental feature (as of 2009-03-22). Some or all Project Gutenberg Mobipocket files may be buggy or may not work altogether.

[Donnageddon]

My initial response is ambivalence. While this may be a boon to those without the time or technical knowledge to transform the books into a format of their choice and design, I have thus far been unimpressed with the ePub versions from them that I have sampled.

I just hope this does not distract them from the absolutely fantastic job they do of transcription of the original sources into HTML and plain text.

But maybe they can learn from experience and improve the outcome of their more specialized formatting.

They will still get donations from me!

[mtravellerh]

I already found one book not working very well (an illustrated book about Darwin where the illustrations didn't work) and I am firmly convinced that automatical conversions can't be perfect (and our "manual" conversions here at MR are far better), but as far as I can tell, the first experiences are quite okay.

[HarryT]

Originally Posted by mtravellerh

No such luck for TOCs. The inline TOCs seem to be working (so far) but no outside one (it would have been quite easy via a simple anchor definition, though).

Could you elaborate on what you mean by this? What is the difference between an "inline" TOC and an "outside" one? Do you mean that there is no "Guide Item" to the TOC?

[mtravellerh]

Originally Posted by HarryT

Could you elaborate on what you mean by this? What is the difference between an "inline" TOC and an "outside" one? Do you mean that there is no "Guide Item" to the TOC?

Yes, exactly. There is an HTML TOC inside the books, but it has not been implemented as a guide item. Inline TOC, for me, is a TOC inside the body of the book, whereas outline means that you are able to call up the TOC via menu in your reader.

[HarryT]

Thanks for the clarification. As long as the TOC is at the start of the book (I hate it when companies like Baen put it at the end!), that's not particularly a problem, since one always has an option to go to the first page.

[mtravellerh]

Originally Posted by HarryT

Thanks for the clarification. As long as the TOC is at the start of the book (I hate it when companies like Baen put it at the end!), that's not particularly a problem, since one always has an option to go to the first page.

You'll have to wade through the blurb, though.

[CommanderROR]

Well, I'll make a completely uneducated guess here...

Amazon (owner of mobipocket) saw the stunt Google and Sony did with their massive ePub library and wanted to do something similar so they went to PG and asked them to provide their library as .mobi files.

This step makes sense in more than one way. On the one hand they have a large library of well-formatted books (not as many as google, but probably less errors) and in addition (an this is where the conspiracy theories come in) they start profiling .mobi as their DRM free PD format (which you are welcome to use on the Kindle) and start widening the gap between .mobi and .azw (or whatever the Kindle file format is) that is their commercial format, available on Amazon only and useable only on the Kindle.

Something to think about perhaps?

[AprilHare]

Originally Posted by CommanderROR

Well, I'll make a completely uneducated guess here...

Amazon (owner of mobipocket) saw the stunt Google and Sony did with their massive ePub library and wanted to do something similar so they went to PG and asked them to provide their library as .mobi files.

This isn't how I see it.
I rather think PG has seen the takeup in the Kindle and the Sony Reader (which now use mobi and epub respectively) and is moving to support them. Maybe the shoddy OCRs for ebooks by Google (hidden behind a commercial website no less) spurred them on somewhat.

[mtravellerh]

Originally Posted by AprilHare

This isn't how I see it.
I rather think PG has seen the takeup in the Kindle and the Sony Reader (which now use mobi and epub respectively) and is moving to support them. Maybe the shoddy OCRs for ebooks by Google (hidden behind a commercial website no less) spurred them on somewhat.

Just a little reminder: The bigger PG Epubs won't work on Sony Reader because the PG Epubs aren't split up to respect the 300kb size limit of the reader. Can that only be considered an accident, then?

Nothing like that in sight as far as Mobis are concerned (works on Kindles, BeBooks, Cybooks, you name it)

[ricdiogo]

Originally Posted by Donnageddon

I just hope this does not distract them from the absolutely fantastic job they do of transcription of the original sources into HTML and plain text.

PG is using a converter running in the background. EPUB files are automatically generated the from the HTML file (if there is one), else from the TXT file, and are meant to be read on memory-impaired devices like mobile phones (hence the lack of images in most of them).

PG is eager to listen to user's complaints/suggestions on how the files show in their reading devices and how can the files be improved.

If you don't have an account please create one here, and post your comments on this discussion page.

You can also join our discussion groups gutvol-d (general discussion) or gutvol-p (programming issues).

Or contact PG's webmaster Marcello Perathoner, who's currently developing the conversion tool, via email or via this wiki page, .

[mtravellerh]

Originally Posted by ricdiogo

PG is using a converter running in the background. EPUB files are automatically generated the from the HTML file (if there is one), else from the TXT file, and are meant to be read on memory-impaired devices like mobile phones (hence the lack of images in most of them).

PG is eager to listen to user's complaints/suggestions on how the files show in their reading devices and how can the files be improved.

If you don't have an account please create one here, and post your comments on this discussion page.

You can also join our discussion groups gutvol-d (general discussion) or gutvol-p (programming issues).

Or contact PG's webmaster Marcello Perathoner, who's currently developing the conversion tool, via email or via this wiki page, .

Well thanks, Ricardo. Do you know anything about this Mobipocket project you'd like to share with us? I am still not sure about the scale, seeing that today's updates don't feature any Mobis at all?

[Hadrien]

Originally Posted by ricdiogo

PG is using a converter running in the background. EPUB files are automatically generated the from the HTML file (if there is one), else from the TXT file, and are meant to be read on memory-impaired devices like mobile phones (hence the lack of images in most of them).

PG is eager to listen to user's complaints/suggestions on how the files show in their reading devices and how can the files be improved.

Well, it's rather the opposite: they're not designed to be read on memory-impaired devices, since you're using a single HTML flow instead of multiple flows.

Overall, the quality is pretty low with such direct conversions, if PG really want to branch into end-formats they should add proper structure to their files first.

[mtravellerh]

Originally Posted by Hadrien

Well, it's rather the opposite: they're not designed to be read on memory-impaired devices, since you're using a single HTML flow instead of multiple flows.

Overall, the quality is pretty low with such direct conversions, if PG really want to branch into end-formats they should add proper structure to their files first.

You are of course right, Hadrien, as I have stated above already (regarding the Epubs) The mobi files, however, look pretty good.

[DaleDe]

Originally Posted by ricdiogo

PG is using a converter running in the background. EPUB files are automatically generated the from the HTML file (if there is one), else from the TXT file, and are meant to be read on memory-impaired devices like mobile phones (hence the lack of images in most of them).

PG is eager to listen to user's complaints/suggestions on how the files show in their reading devices and how can the files be improved.

If you don't have an account please create one here, and post your comments on this discussion page.

You can also join our discussion groups gutvol-d (general discussion) or gutvol-p (programming issues).

Or contact PG's webmaster Marcello Perathoner, who's currently developing the conversion tool, via email or via this wiki page, .

One good fix would be to adopt Calibre as your conversion tool. It does a good job of producing both ePUB and now mobi as well as several others.

Dale