- From: John Foliot <john@foliot.ca>
- Date: Fri, 23 Jan 2015 18:01:30 -0800
- To: "'Wayne Dick'" <waynedick@knowbility.org>, "'Duff Johnson'" <duff@duff-johnson.com>
- Cc: <w3c-wai-ig@w3.org>
Wayne Dick wrote: > > My comment was directed at the original question from Rachel. She is > from a university setting looking at media for instructional materials > in the United States. That is a very different issue from general > accessibility of PDF. > > In general, PDF requires more intervention to create an accommodation > than other file formats, and that is usually a lot more intervention. This is especially true when the content source is vague or lost (i.e. the source may be a scanned document). OCR is good, and getting better, but OCR suffers from the 'garbage-in, garbage-out' problem, and fuzzy scans, small font-faces and other artifact issues all get in the way of a fast-and-easy solution. As a result, many of the PDFs that originate that way can be a heck of a lot of work to remediate. If I were king of a University, I'd be steering that University to EPub, as IBM is doing internally (see Phill's email), as it does a very good job of accessibly addressing many of the "requirements" of PDF publication (real or imagined). > So, in Rachel's case, PDF is generally a bad bet. It is hard to teach > faculty how to make accessibly marked up PDF. Accessibility support is > spotty. This translates into higher accommodation costs for the > disabled student services center. The "teach faculty" problem is driven, not so much due to the complexity, but rather the lack of time or interest to go the extra mile (sic), and many who have spent any real time in academia will have heard the "academic freedom" refrain at least once or thrice. (It sucks, but it is what it is.) Educating content creators to start with well structured word documents will, as Duff has noted, go a long way in mitigating both the effort and cost, but it *does* take a concerted effort and 'program' to manage and corral accessible PDF creation in large enterprises, be they academia or big business (I've worn both hats, same old same old). > I don't know about other contexts, but I do know academic needs. I > think John Foliot understands these issues. I actually wasn't > commenting on the accessibility of PDF, just whether it was worth while > using it as a dominant medium in an instructional support web > environment. PDFs are victims of the 80/20 rule, where they "work" for the majority, and then start to fail after that (Note: I am not suggesting that 20% of all users cannot access all PDFs, but the rule does apply none-the-less). For me, the final decision on whether or not to use PDF is contextual: there are many things to consider, including source, content author skills, intended audience, intended use (I mean, if one of the primary reasons for posting a PDF is to provide a visually rich, printable document, then it may not be accessible to all for "reading", but even vision challenged users can save that file and send it to a printer) so context is important. After all, Pantone colors aren't very useful on the web, or for vision impaired users either, but they *do* have an enormous value as well, so step back folks, look at the bigger picture. > If I sounded strident I didn't mean to be iconoclastic. I just thought > that an academic coordinator for accessibility could use a warning > conderning the effectiveness of PDF for instructional materials at a > university Agreed, in a Uni context, I would look to ensure that a full policy included training opportunities, responsibility and resource consideration$, tracking and testing/verification capability, etc. It may not be cheap to implement a program like that, but failing to do so could cost more (IMHO). As a general suggestion I would counsel to try and avoid using PDF whenever possible, but personally I fall short of "forbidding" or ostracizing them outright. JF
Received on Saturday, 24 January 2015 02:02:02 UTC