Best "Plain Text Conversions" Settings for .MOBI?


When I created .MOBI files previously (hand-coding HTML in JEdit) I remember reading that I needed to convert three dots (…) into ellipses, and to convert curly quotes into smart quotes (before importing the text from Text Wrangler into JEdit).

Thus far, I believe I’ve been doing the opposite (by checking the first and third boxes in the “Transformations” section of the .MOBI Compile Settings), and the resulting eBook looks great.

I’ve tried unchecking the boxes and compiling, and it seems to turn out all right as well. Can anyone advise as to which is the safest way to do it (that is, the way that is most likely to not have conversion issues across different Kindle models)?

Thanks in advance!


Even the first Kindle supported UTF-8 encoded HTML files, which is what Scrivener produces. UTF-8 does not suffer from some of the problems that existed with older character set encodings which would cause special characters like typographic punctuation to appear as question marks in browsers that didn’t support the codepages. This is what caused the entity system to become the recommended way of printing special characters. So in the HTML code it would be “ instead of “. With UTF-8 it is not necessary to use entity escapes, so long as the display software is capable of addressing UTF-8.

Technical gobble-gobble aside: you can use fancy punctuation in your documents with the Kindle as your target device (and most other modern reader hardware & software).

Excellent to know…to be honest I had felt a little bit guilty about using Scrivener to export the .MOBI file (after hand coding html, I felt like I was cheating (but…mostly I was just incredibly thankful for Scrivener!)) At the back of my head I had the nagging feeling that I would be offering my customers a subpar .MOBI file…but I’m not super tech minded (obviously), so what do I know?

Sooo…does that mean that it doesn’t matter (from a usability standpoint) what I choose in the Transformations section? And so the only question, then, is one of aesthetics?

Yeah, it’s pretty much just a matter of aesthetics. Something you might want to know, if you like to make sure the output is clean, is an option in the KindleGen pane that will let you export the full source that is used to generate the .mobi file. You can get in there and make fine-grained tweaks to the code if you wish, then run KindleGen against the OPF file, or load it in Kindle Previewer to generate a .mobi file. In general I would say the output is pretty good though.

I think Three Press Consulting gave a very fair review of Scrivener’s e-book output (they reviewed the ePub output, but Kindle output is almost identical in terms of HTML etc): … scrivener/

Their main gripe was that Scrivener outputs too many CSS files - one for each HTML file - which is true; a result of us using Apple’s HTML exporters rather than hand-coding our own and thus culling the CSS from each, which doesn’t make it possible to combine the CSS into a single file. But other than that, output should generally be good.

All the best,

Awesome…thanks so much, again. That makes me feel much better about using Scrivener for the entire process (I’m constantly finding reason to love, and recommend it!)