Bobulate


Posts Tagged ‘pdf’

PDF Readers near you

Wednesday, September 15th, 2010

Do you get this as well? A PDF delivered along with a message that you can use Adobe Acrobat Reader (r) to open that file?

PDF is a (relatively) open standard. It is an ISO standard (19005-1), for one thing. This also means that there are alternate implementations of the standard. And you might have good reasons to avoid the Adobe implementation. For instance the number of exploits against their implementation, or because it doesn’t run on your hardware / software platform. Most of the time I’m at a computer which is perfectly capable of dealing with PDF files through a Free (both as in speech and in beer) PDF reader, and the "download Reader" just strikes me as weird. I get these PDF files from travel agencies, hosting providers, financial advisors and local governments. I often reply asking them to update the text accompanying the file to say something like "You need a PDF reader to view this file. Get a free one or use something else." The PDF Readers (.org) site is a good place to point people and organizations. The site points out the available options and how to get a Free PDF reader.

The Free Software Foundation Europe (FSFE) has started a campaign to promote the use of Free PDF readers by local governments. The idea is to point out where your local government is pointing only to the proprietary software solution and to get them to adapt the text to offer more choice. There’s a contest involved, as well.

As far as the PDF readers site goes, it points to muPDF; that’s one I would personally avoid for semi-technical reasons: the code is terrible and utterly undocumented. Maybe the application works — but it’s not something that satisfies my code-readability test. Not one of the files has a license header, although the thing as a whole is licensed under the GPL version 3 or a commercial license.

Turning to the more popular — or rather, the recommended readers on Free Software operating systems — selections, there’s Okular and Evince. I use both of them fairly regularly — Evince is my fallback on OpenSolaris during those times that I’m compiling KDE. Following the links from PDF Readers to the two web pages for Okular and Evince shows a pretty big difference: the latter is focused on packages and contains a link to a how-to-compile page, while the former is all about building the software from scratch with a comment that there’s probably packages available. Neither of these strike me as a particularly good user experience. I wonder if there’s any feasible (technically and privacy-preserving) way to detect the OS so as to improve the download suggestions.