Quote:
Originally Posted by Gregg Bell
This one of yours has been wearing on my mind a bit. The way I've often done ebooks is to start in Word, save as html (in Word still),then put the html in a text editor and clean it up there before putting it in Sigil. Now you're saying that I'll get a cleaner save with Writer2Epub. But what am I "losing" if I save (now it would be in LibreOffice and an odt-->html) as an html? Writer2Epub focus on cleanly exporting "the actual words." Well, what is the odt-->html focusing on? Thanks.
|
preserving all the code. Every time you fiddle with the font size/type, margins, or anything else whatsoever, and I do mean anything, it creates a set of tags to change the meaning of just that little chunk. Then backspacing a word, adding in new words, cutting/pasting stuff around, divides it up even more till you have an absolute soup of a mess, which you don't see till you try editing in an epub editor. Ideally, writers should use nothing but styles, and then a straightforward conversion would be nice and neat. writer2epub is designed to filter out most of that junk, whereas odt-->EPUB conversion will try to duplicate the look and feel exactly.
By cracking open a book in Sigil/calibre, you can often see a horrible mess of span tags containing font sizes/families and margins, bold/italics turned into yet more span tags, and worse stuff. Oftentimes each word/phrase in a paragraph will have the same styles individually applied to that word.
Obviously, your mileage will vary depending on the book. And MS Word (used without styles) is the worst offender. But writer2epub for LibreOffice and Toxaris' addon for Word are supposed to occupy a sane middle ground between leaving in the cruft and going the Smashwords Nuclear Method then going in and reapplying the formatting.
Really, it all depends on how clean your editing process is, I suppose.