..changes in the software, a computer crash, a new computer, and more changes in the software…and issues right now plaguing everybody who works with e-books, have kept Jane busy through tonight just trying to format Chernevog so that it looks right in ALL displays. You get Kindle right—and Nook refuses to display correctly. You get Nook—you guessed it. She’s done nothing from 7 AM to 11 PM and beyond, sometimes to 2 AM…and the beast is full of dodges, hidden code, and dastardly schemes to prevent files from working — just WORKING sanely. The covers on which she has labored long and hard—want to stretch. Or shrink. The software folk are paying no attention to the long-held book industry standard for cover size. Oh, no. Amazon wants one size. Nook wants another. Calibre, the conversion software, produces yet another, sideways-stretched. Mobipocket creator, deliberately bought and orphaned by Amazon, is still reasonably good. Sigil is hiding code somewhere. I’ll tell you, it is NOT just push-a-button and convert. These companies are fighting each other…I strongly suspect…doh!…and the idea of giving the creator control with easy, visible toggles is not on the horizon. Sigh.
So when you do see the new things, note very long hours by Jane and by Lynn working on these conversions. Think of them, please. I wish I could do something to help—but at this level of computer chicanery, all I can do is play librarian, searching the internet for answers to very specifically posed questions. 9 times out of 10, they’re ahead of them. Just rarely I can find something useful. But alas…not often enough.
To be crude … that s..s!!
I bypassed all the conversion tools and went to very vanilla HTML.
I hand-code the html, but it’s not all that time consuming.
It’s been much much easier since then, and I have far more control.
I clean up my WP generated html before converting to epub as well. My code is very clean going into Calibre. The older books, not so much. I’m smarter now, but cleaning up and putting the new covers in means backtracking through several layers of learning curve on some of the files.
And I like pretty chapter and section headers, fonts for letters and notes within the file, which means formatting and embedded fonts. I use Calibre to convert the html to epub, create the TOC and opf files, etc. No matter how careful I am with the initial html, Calibre makes a very messy style.css file. I tried at one point to clean it up, but changes that shouldn’t make any difference, and didn’t to any other reader, makes ADE give up and display the expletive deleted thing as a simple txt file. (ADE s/b blasted off the internet, IMO!)So, when I was still trying to sort everything out and had a dozen books to convert…I left it messy.
And then there’s the conversion to mobi. I initially used mobi pocket creator because it gave all the requisite parts for a kindle file very nicely from my html…but now it doesn’t work with the machine and besides, I’d like to simplify the process. Getting into the opf file and making the changes I outlined on my blog has made that possible to do using a good epub file and kindlegen…but that has all taken time and the combined efforts of Lynn and myself.
Sigh…back to editing the next ER book…
I have only dealt with Sigil and Calibre to get my own writing onto my Kindle to see how putting together an e-book works, and it is frustrating. Sigil constantly changes your code for no good reason, and Calibre is the same. I cannot imagine trying to make a book that works across all platforms.
Calibre and Sigil always have their reasons…part of what we’re doing is trying to figure out those reasons so we can simplify the files and still have them work on as many readers as possible. The real devil is Adobe Digital Editions. You can have a file with lots of nice formatting and embedded fonts that works perfectly in everything else, and that stupid thing will display it as a simple txt file.
At one point, I lost all formatting in ADE…turned out it was Sigil not liking the M/i> that Calibre was still using at the time and putting in it’s own ital CSS code. Tracking down that one was fun…
I’ve since learned to strip that out of the html file with a css font-style command before doing the conversion, which helps a lot.
Cleanup of the WP generated html is vital. I don’t know what you do with a file created by Word. Those are the messiest html files I’ve ever seen! (shudder)
We’ve had less trouble in the past…but the software is a moving target, then to discover that the new computer has to be converted from hyperthreading to not to manage the software—it’s maddening. OTOH, if you get a choice involving a 2 hard drive laptop with one a flash drive type—it’s a good thing.
From Jane…a moment ago. “No, no, no, no, no—” And: “What on earth?”
SOmething else has gone wrong with Calibre. In another platform.
What’s this less trouble in the past, Pocahontas? Why do you think it takes me so long to gird my loins to take on the monster? 😀
And what it went wrong in—of course—was Adobe reader. They don’t play nicely with anyone. But she believes she’s got it.
The writing of a book is a shaggy process with many moves, changes, and amendments and insertions. All this to-do can leave code behind, because word processing programs want to keep notes on what was done. One first step is the conversion to HTML (soon HTML5, apparently) —which can drag along some of these bits. Unfortunately we haven’t found any cleanup utility that doesn’t ‘clean up’ things it shouldn’t touch. And writing is not a straightforward dictation into a machine. Every line may have 3-4 versions, backups, insertions. The whole line may have been moved to the end of the paragraph. The paragraph may have been moved up in the text…This goes on thousands and thousands of times. So a tiny smidgen of code survives—and it causes these reader programs to insert a font change or a line width change or a size change…it’s two worlds who do not play well together. Printing out and scanning in again produces yet another array of artifacts.
It’s maddening. A few toggles or a ‘reveal codes’ option would make life ever so much easier. Word Perfect, thank goodness, has ‘reveal codes,’ and so does ‘Namo Web Editor,’ but it’s still a nest of snakes, some venomous.
Pooh!! It’s easy to clean up the html code. And the entire process is getting easier. You just have to learn what’s important. What I’m fighting now is my own learning curve, which didn’t involve as much cleanup in html before converting. I’ve finally gone back to the WPD file and am starting over. I think there will be much less blue smoke about the house. 😀 (Ah, the wisdom of staying up til gawd awful fighting a file….)
I’ve seen what the various word processors export as HTML. It’s not pretty. It requires some (or a lot) of hand-cleaning the HTML and CSS code.
I see that EPUB 3 is supposed to be based on HTML5 and CSS3. That’s good, actually. But I’m pretty sure EPUB 3 is not yet supported (output) by the software.
I know dang sure word processors aren’t exporting standards-compliant HTML4.01 or XHTML1.0 or 1.1, or CSS2, so, well, it could be a while.
Meanwhile, Jane and Lynn and others, and those just learning EPUB like me, will get to jump through hoops and learn that it is largely hand-editing the blasted stuff.
At least I have a fair amount of experience with the HTML/XHTML and CSS side of it and cleanup of the exported word processor output. So far, it looks like I have more of a head start than I’d thought I would. Yay.
But I’m only in Chapter 3 of Liz Castro’s book, EPUB Straight to the Point. The only fault of the book seems to be it presumes you’re using either Word or InDesign, though she recognizes people will be using others. The book is only a little over a year old, and already it’s getting slightly dated. I expect she’ll put out a new edition for EPUB 3. She’s been doing web design books for many years now. Skilled lady.
The next two chapters get into the real deal, so I should have a slightly better idea what I’m in for after reading those. I decided to put this as first priority for awhile.
WP does a pretty good job if you use the heading macros and such for your chapter titles religiously. The problem is, Calibre “batches” every formatting nuance for a paragraph rather than using simple “ital” “bold” “indent” and such. It’s very annoying to try and sort them out. Lynn uses simpler formatting than I do and she’s created a very pretty, easy to understand, .css file. I was trying to do something similar to Chernevog when the whole ADE thing blew up in my face. It was looking fine on Sigil, Calibre and my nook…then I brought it up in ADE…and it was txt. BLEH! So then, I began trying to reproduce the effect to find out what code change did it…which was a nightmare of make a change, save down, bring up in ADE…Finally gave up in order to get the book out. It’s fine as is, but not what I’d like to have underneath.
But then, I’m one of those people who paints the back of furniture designed to go against a wall….
The #1 bug in ADE is that it completely ignores the whole css file if it finds any error in it – like a missed semicolon – grrr! The only protection is the css validator from W3C: http://jigsaw.w3.org/css-validator/ – which is in addition to the ThreePress ePub validator (which doesn’t check the css files). On more step in the work flow, even if it’s only when ADE borks on you.
There is a big long thread on MobileRead about ADE bugs… but it’s used on Sony, Nook, Kobo, and others.
My own workflow uses Libre Office to convert an rtf file to a web page, then HTML-tidy to convert to xhtml and collect all the font styles into a style section. Then I delete that and put in a link to my own standard css file. I have a bunch of shell files that do editing for me, which helps a lot, but it’s not a high-throughput process. And that’s just for the ePub files; I’ve barely started on mobi and K-fire files.
Supposedly Sigil validates before saving…it certainly won’t let you get off a page if you’ve made an error, but I’ll look into that basic reality. I’m pretty careful, but that could be it. The other great thing is if you get a cap wrong in a file name. ADE has fits about that, too. You have no idea what I went through tracking that one down.
If you’ve got a good solid epub, the transition to Kindle is pretty easy. See those additions to the opf I mentioned on my page. They don’t seem to interfere with the epub (even ADE accepts them!) and they’re all you need to make the “go to” in the kindle work as well as giving you the thumbnail for the library displays.
I use Word and I always do reveal codes. I made a manuscript template that has all the formatting the way I like it, all the margins and tabs preset, special fonts, little graphic things, page numbers, etc. I’m working on some other Word files from Shadow Unit (with permission of authors!) to copy files to Word and try formatting them to mobi so they look right.
Jane’s trials and tribulations brought to mind that scene in the old Disney version of 20,000 Leagues Under The Sea (The one with Kirk Douglas and James Mason) where the Nautilus was attacked by the giant squid. . . . .http://www.youtube.com/watch?v=0qtWzbDgIQE&feature=related.
Oh, yeah…squid. Definitely. Great movie.
Going from “scratch” that’s what I do now, but these books started with every kind of file from Volkswriter to ocr to wpd. And they’ve been through multiple stabs at formatting to find the look I like. Even so, each one dictates its own look…certainly each TOC has different needs. I think my favorite, and what I’ll probably go with, is what I’ve done for the GroundTies books, which is sections with titles and subchapters beneath that in a string. It allows for a condensed, single-page TOC with the section titles to give a hint to where you are in the book.
But I’ve experimented with all sorts of looks for the TOC and chapter titles and getting the auto generated Tocs to contain the necessary information. I know…I make my own life difficult, but I’m trying to maintain some of the details that I always enjoy in a DT book…
giant squid = tons of calamari! Yum!
Jane, when you get done with your html, how about coming over and helping me with mine? I still have no idea how to construct a css, or even why I would use one, if it’s just a small website.
You have so much more patience than I, it’s amazing how you tackle a problem and then seem to have it solved much more quickly than I could.
Obsessive/compulsive, I fear. Sigh… The weirdest thing with all this is how clean my file will be going into Calibre and what craziness comes out the other end! 😀 It’s those tiny variation that ADE finds and decides aren’t viable that will drive me nuts!
For those frustrating times when someone or something really needs a smackdown,
http://www.renblades.com/Spiked%20Mace.jpg
We call it the “Science Stick”, for when an edumacation is needed. It’s an old WWI trench mace.
Makes me want to load MMVI on my machine…I need to shoot some goblins….
Jane:
I don’t know if this would help (or if you already know about this) but there’s a css validator available from W3C (http://jigsaw.w3.org/css-validator) that’s helped me sort out css problems in the past.
Andy
Yeah, the “ignore everything if there’s a single error” behavior sounds familiar; I helped put together a web site for a local candidate, who used a webhost specifically oriented for political campaigns, nationbuilder dotcom. The nationbuilder tools were like that, except that I have NO idea what software tool they used to filter the HTML. For all my hardware and software knowledge, I’m a rank beginner at HTML, this was my first venture as semipro (and they couldn’t find/afford a real expert), but I learn quickly and had context to copy or adapt from. I once spent 15 hours cleaning up the (what I now know to be auto-generated) code — and incorporated changes the candidate wanted — uploaded it to the site, and saw it turned back into (more or less) the original mess in less than an eyeblink. Argh! Actual competance, it seems, is not allowed, just let the “wizard do it for you! Feh. It misplaced the link half of all of my anchor/tag pairs. Figuring out how to make the necessary changes “stick” was the real challenge. And it wouldn’t accept plain (perfectly valid) HTML, you *had* to include specific (otherwise useless or redundant) js formatting tags in every command, or it would — you guessed it — remove the whole “font=” command, or whatever you were trying to do.
Best help I got was HTML for dummies, that lists the basic code for positioning a picture, defining a page, getting caps, columnizing text, etc, all those wonderful things.
The webeditor I like is NAMO, which gives you a nice wizzywig (what you see is what you get) type it interface on one tab, the actual code on another, and Preview on the third. With that and Dummies it’s a pretty sane way to operate.
I would also recommend books by Elizabeth Castro:
Castro, Elizabeth. HTML, XHTML, and CSS Visual QuickStart Guide, 6th Ed.
Castro, Elizabeth. HTML5 and CSS3 Visual QuickStart Guide.
Castro, Elizabeth. EPUB Straight to the Point. — I’m reading this one now.
Also, if you already know some HTML4 or XHTML1 and want to get up to speed on HTML5:
Introducing HTML5 (2nd Edition) by Bruce Lawson and Remy Sharp.
@ Joe — The reason to use CSS, either in a style block in the head of your HTML page, or as a separate file, is consistency and ease of changing nearly all your presentation formatting in one place, rather than each instance directly in the file, or in umpteen pages or chapters. It separates content (the meaning and information) from style or presentation (how it looks, what makes it purty). If I set up my stylesheet(s) well and use very plain-vanilla HTML with class names and ID’s, then in theory, I can completely change the look just by providing a different stylesheet (CSS file). Yes, it takes some getting used to, and the print versus screen difference can be hard to wrap your head around at times, but it’s worth it.
Avoid, avoid, get rid of, the font tag in HTML. Bad, bad, old, old, even though many tools still put it in. Use CSS instead.
‘Fraid I can’t help! I use basic, would work 15 years ago, html in my base file. I then let Calibre create the CSS file. I understand how it works and can really clean it up, and I could create it first…but it’s easier for me this way. I honestly don’t know html5 from html…1? Whatever. I don’t need anything fancy at this point, and if I don’t need it, I don’t figure it out! 😀 😀 😀 I learned html in the first place by examining the code from other sites. 😀 Created a pretty danged complex, highly graphics-oriented, interlinked web page that way!