Well, Amazon has announced the Kindle DX today. It’s got a near-full-page sized display and native PDF support. So, no longer is there a need to create a smaller e-book screen sized version of our scientific publications as I mentioned here. (Thanks Amazon!) The price for the DX is not too much higher than the Kindle 2 either, so it seems pretty attractive to me. After it’s released we’ll have to see if the PDF support is as good as Amazon claims.
With all the media coverage of the Kindle and Kindle 2 I started looking into eBook readers again. It’s been a few years since I’ve taken a good look and the e-Ink display has been in and out of the scientific and popular press for about 10 years now, so I figured that there may be something good in the Kindle or other readers out there today. From what I can tell the hardware is similar on most eBook readers, including Kindle. The difference is in the software and supported file formats.
The biggest use for me would be reading scientific journals and magazines as well as individual PDF based articles downloaded from the IEEE and ACM libraries. Kindle doesn’t natively support PDF but some other readers do. I started to wonder why this was and began to poke around the Internet looking for a conversion solution to get a PDF onto a Kindle for free. I know there is the “experimental” PDF conversion, but people claim that it doesn’t work so well for all but the most basic PDF files. I found a few solutions, but the majority of people who don’t use Amazon’s converter use a Python solution that converts PDF pages to PNG files, does some dilation to enhance the text, re-PDFs the PNGs and then converts that PDF to a Kindle compatible format. Yikes. I tried that on an issue of Communications of the ACM and the text was so small it was unreadable. Which actually makes sense when you figure that the conversion process is shrinking the already small 10-11 pt font used in the magazine by at least 50% to fit on a 6″ diagonal Kindle screen.
OK, so Kindle or any small screen eBook reader is not going to be useful for a converted magazine or article using the Python script PNG method. Bummer. Then I got to thinking, why should I have to convert these files. They’re digital, repagination is relatively simple for the creator, we submit our articles either in the source format (.doc or .tex) or in pre-formatted PDF. We should be publishing our work in both full 8.5×11″ size and a smaller 3.25×4.5″ eBook friendly size. Both should be made available via the IEEE and ACM libraries, formatted and submitted by the authors. Even better would be a way to publish directly to the Kindle or other eBook formats and keep any table and image formatting.
I did a test with one of my old articles in LaTeX and by changing a few lines dealing with page size in the header file I was able to compile an eBook sized document. It took all of 30 seconds (once I figured out the appropriate LaTeX commands). You can see the results here. craighead_cts2008_ebook Now that PDF still has to go through the conversion process with the Python script, but the images do not need to be scaled down, so the resulting PDF contains an easily readable image of each page which can be converted to the appropriate format for your eBook.
I’m going to start making my articles available in an eBook friendly format and you should too.