We no longer support .epub versions of our OpenStax books. Only a small percentage (around 1 percent) of OpenStax readers reported using the .epub format, and maintaing these versions is costly; instead, we focus on providing our online web view, PDF, iBooks, Amazon, and print versions. You can see what formats are available for your subject at OpenStax.org."
I call bullshit. EPUB is basically a self-contained web site with a built-in index. If the text is maintained as an EPUB to begin with, producing a PDF is functionally identical to rendering it on an EPUB reader with a fixed page size and font.
They provide an online web view. The margin of difference between that and EPUB is building the metadata file and ZIPping it all into a single file. Or it's just an online PDF reader. Besides that, iBooks is based on EPUB, and older Kindle files are based on MobiPocket, which is based on OEBPS, the precursor of EPUB. Amazon has gratis software to convert from EPUB to Kindle file format. The print versions are probably functionally identical to PDF, with printing and cutting margins added to the page size.
If they maintained the book source in EPUB, all other formats that they do provide are essentially free, and can be maintained with a conversion program, a shell script, and cron job.
The real reason is that a lot of the available EPUB readers do not render web pages consistently, and they don't want to test Adobe Digital Editions, Calibre, etc.
They used to do it and made an intentional decision to stop. They say it’s because it was costly. You come along and say they’re lying (meanwhile acknowledging the testing burden). I don’t get it. You think this non-profit, dedicated to the widespread distribution of their content, has some unspoken nefarious reason to deny you ePub versions?
I am unsatisfied with their explanation for discontinuing EPUB. "Too costly" is a subjective judgment, and they did not reveal the objective measurements or methods they used to reach it. A pack of gum is costly if you already spent all your money on something else.
With the knowledge of what they do continue to support, I suspect that some of the costs they are paying for maintenance are the result of poor architectural decisions. If all book formats had a common source format, supporting any particular book format would be a matter of maintaining a compiler/converter for it, which could be a large initial cost followed by a lower recurring maintenance cost, shared across all books in the system. However, open sourced, community maintained programs are readily available, along with proprietary commercial offerings with paid support.
I know that the difference between raw web site and EPUB is some OEBPS metadata and a ZIP program. The difference between EPUB and iBook is a conversion program maintained by Apple. The difference between EPUB and Kindle is a conversion program maintained by Amazon. The difference between EPUB and PDF is hairier technically, but conceptually it's just non-reflowable fixed page sizes and fonts, which are theoretically handled by CSS "@media print" directives. The difference between PDF and print depends on the printer, but for self-publishing it's mainly just getting the page margins correct between the digital and print PDF versions.
Testing is a branding issue, not a technical burden. You don't actually have to test Adobe Digital Editions support, if you warn people on download that some EPUB readers are not fully compliant with the EPUB spec, and therefore will not render correctly. It's the IE6 problem all over again. If you support only the readers that are currently popular, there is no incentive for their maintainers to move any closer to a standard that would make maintenance easier for you in the long run. You provide the book, not the reading experience. That is the responsibility of the reader.
Mainly, I'm just pissed that they axed EPUB, but still do iBook and Kindle. That is implicitly supporting closed, proprietary formats (or embraced and extended formats) rather than open standards. We have already been there once.
> Mainly, I'm just pissed that they axed EPUB, but still do iBook and Kindle. That is implicitly supporting closed, proprietary formats (or embraced and extended formats) rather than open standards. We have already been there once.
Yes, exactly, that was my reason for asking to begin with.
But I can buy into their explanation that amazon/iapples are much more downloaded as the number of devices are probably a couple of magnitudes larger.
But still sad a project with "open" in it's name mainly seem to support proprietary formats.
Hahaha, look closely at the PDF and online version (just look at the math). There are a ton of corner cases which EPUB and especially the EPUB readers do not support. It's definitely not convert into EPUB once and then make books out of it with shell scripts. This is a very wishful thinking of the EPUB standard (which is not a bad standard but it's not that overall solution you are stating here).
> The real reason is that a lot of the available EPUB readers do not render web pages consistently, and they don't want to test Adobe Digital Editions, Calibre, etc.
because that's necessary to maintain an acceptable epub offering, and is costly?
One of the reasons is that the books are often used in printed format in classrooms and it's important that the page numbers of the ebook match what the teacher assigns from the printed copy.
With epub format the page # will vary depending on font size.
There's probably other reasons as well, but that's one that I've heard mentioned.