(1) is genuine (but then many websites assume constant width, so it’s not PDF-exclusive issue), but (2) and (3) sound totally made-up to me. Browsers have had far more security vulnerabilities, and use far more CPU/memory than PDF readers.
Indeed, I’ve been infected by PDF-based viruses more than once. Updating Acrobat and turning off JavaScript in PDFs isn’t enough to keep you safe, either; I finally added NoScript to Firefox in order to prevent any PDFs from being displayed without an extra enabling click, so that only PDFs I trust are ever downloaded.
Of course, this has little relevance to scientific papers: the PDFs that you need to worry about are the ones that you never intended to download in the first place, that are downloaded in the background via JavaScript or an iframe embedded in an ad on a random webpage. (I once caught one from Kaj Sotala’s LiveJournal page, for example… just visiting the page was enough to infect my machine.)
Nearly echoing FAWS, a browser alone will have less CPU/memory usage than a browser+a PDF reader. More importantly, there is no delay to load the PDF viewer when visiting an HTML page, where there is for PDFs.
(1) is genuine (but then many websites assume constant width, so it’s not PDF-exclusive issue), but (2) and (3) sound totally made-up to me. Browsers have had far more security vulnerabilities, and use far more CPU/memory than PDF readers.
PDF viruses exist.
Indeed, I’ve been infected by PDF-based viruses more than once. Updating Acrobat and turning off JavaScript in PDFs isn’t enough to keep you safe, either; I finally added NoScript to Firefox in order to prevent any PDFs from being displayed without an extra enabling click, so that only PDFs I trust are ever downloaded.
Of course, this has little relevance to scientific papers: the PDFs that you need to worry about are the ones that you never intended to download in the first place, that are downloaded in the background via JavaScript or an iframe embedded in an ad on a random webpage. (I once caught one from Kaj Sotala’s LiveJournal page, for example… just visiting the page was enough to infect my machine.)
But a browser alone will have fewer vulnerabilities (and probably use less resources) than a browser + a PDF reader.
Nearly echoing FAWS, a browser alone will have less CPU/memory usage than a browser+a PDF reader. More importantly, there is no delay to load the PDF viewer when visiting an HTML page, where there is for PDFs.