Today I began a rather large project of updating my Time Management e-book to have considerable additional content, and to make it available in ePub format as well as PDF and Kindle formats (mobi). Since publishing in ePub format is a new experience for me, I wanted to ensure that I have a handle on how to do it before undertaking the massive update. In particular, I wanted to see how well it handled images and how good/bad they looked on an iPhone (using the Stanza iPhone app), so I would know what sort of challenges lay ahead for me in dealing with images. Good images can help make a good reading experience, and bad ones can ruin it.
I had built the initial version of my e-book using several different tools, all of which relied on having the document available in either MS Word or HTML format. So, my initial test of dealing with ePub was to use the free
Calibre tool to convert the HTML source of the original document to ePub format.
But, before attempting to do that, I wanted to make sure I had a clean HTML file. By clean, I mean one that is not cluttered with a lot of bulky CSS that is treated more like inline HTML than like CSS. Unfortunately, the HTML file I had on hand was "dirty" in that it had been created by MS Word, which is notorious for cluttering its HTML output with this weird sort of misuse of CSS.
So, I tried to find a way to generate clean HTML. At first I tried OpenOffice.org Writer, letting it open the original DOC file and then selecting SAVE AS HTML. Every time I tried, it locked up and I had to Force Quit the application. (I am using my iMac for all of this, so far.) I tried this several times, with no success.
Then, I decided to try to have Apple's Pages program save it as HTML. I opened the DOC file in Pages, and selected EXPORT, and the options did not include the ability to save as HTML. What the hell, Apple?
Not being one to give up, I imported the file to Google Docs, and attempted the Export as HTML. It saved a zipped file that would not open.
Now I am getting pissed. So, I decide to try Zoho Writer, as I had never tried any of the Zoho programs before, and figured this would be a good time to check at least one of them out. I imported the DOC file into Zoho Writer and had it export as HTML. It did so quickly, and the HTML didn't look too shabby. It did not use any CSS, but the code was clean otherwise. The images were referenced as links to images stored on Zoho's servers, though. I thought this might be problematic, but decided to load the HTML file into Calibre and have it output an ePub file. It did so, and I uploaded the file to my server, and then used Stanza to download it.
Bummer. The images did not show up. Just little image icons showed in the place of images.
At this point I gave up on my insistence in having "clean HTML." Screw it. I could work with the MS Word generated HTML if it would help overcome these other problems.
Ah, but Calibre did not give me images in the ePub file it generated from the MS Word-generated HTML file, either.
Then I read that Calibre could convert a OpenOffice.org ODT formatted file to ePub format. So, I opened the DOC file in OpenOffice, converted it to ODT format, and tried that. Again, no luck in saving the images.
I did finally find something that works, though. That is to take a prc file generated by MobiCreator and let Calibre convert it. That worked beautifully. A few of my images are going to need some work before they look acceptable on the iPhone, but now I know that it is all doable. And, the nice thing is: I have to use MobiCreator to create the Kindle formatted e-book anyway.
So, now I can write my e-book using MS Word (or one of the other programs that save in Word format), let MobiCreator generate its prc file from that, and then let Calibre generate its ePub format from there. Sweet.
The only downside I know of from having to do it this way is that MobiCreator does not handle Tables well. Although ePub does handle Tables well, it will not have a chance to do so when the document information is passed to it by a MobiCreator prc file. This is a disappointment because there are a few places where I could make good use of Tables in the e-book. I will now have to convert those to images (as I did with the previous version) to get them to look okay in all versions. I would prefer to not have to do that, since the images do not look as nice as the tables.
May take me a couple of weeks to add the content I want to add and to organize it all properly, but I am relieved to know that it is all doable, even if I do have to compromise on the use of tables.
Comments [0]