Hmm … back in the days of adobe acrobat pdfs were web cancer, but these days they seem pretty OK. Being able to convert on command line between docx/odt, html, markdown and pdf is pretty useful.
If pdf UX on smartphones is not great, is that really a pdf problem or a viewer issue?
PDF Viewers can definitely do a lot of things with PDFs, e.g.:
auto-trim the margins
if it’s a multi-column PDF, put the columns beneath each other
switch to landscape mode
try to extract the text and reflow it.
By now, I’ve seen all of these features somewhere, but I’ve never seen them applied automatically and so it always requires a bit of faffing around until I have a reasonable viewing experience. And in particular images are often a problem.
What I could maybe see as a solution for this, would be a “Mobile First”-approach like in UX design, i.e. we would make PDFs that are not suitable for full pages, but rather for quarter pages (so that they fit on a phone screen), and then on the desktop, you should be able to either just piece them together (two-column seems to be a popular layout for scientific papers already), or reflow them with less trouble, because the single-column layout makes it easy to figure out where certain images belong.
Now we just need the whole industry to switch over, which of course should be no problem.
The problem is that pdf wasn’t really designed to be modified in this way, if at all–at least according to my understanding. These problems would be alleviated if people switched to a different, more modular format, e.g. epub, and improved that rather than continue to use pdf in this way.
Edit: To add, pdfs are great for printing, but very annoying to work with on desktop–Windows especially–or edit.
Hmm … back in the days of adobe acrobat pdfs were web cancer, but these days they seem pretty OK. Being able to convert on command line between docx/odt, html, markdown and pdf is pretty useful.
If pdf UX on smartphones is not great, is that really a pdf problem or a viewer issue?
PDF Viewers can definitely do a lot of things with PDFs, e.g.:
By now, I’ve seen all of these features somewhere, but I’ve never seen them applied automatically and so it always requires a bit of faffing around until I have a reasonable viewing experience. And in particular images are often a problem.
What I could maybe see as a solution for this, would be a “Mobile First”-approach like in UX design, i.e. we would make PDFs that are not suitable for full pages, but rather for quarter pages (so that they fit on a phone screen), and then on the desktop, you should be able to either just piece them together (two-column seems to be a popular layout for scientific papers already), or reflow them with less trouble, because the single-column layout makes it easy to figure out where certain images belong.
Now we just need the whole industry to switch over, which of course should be no problem.
The problem is that pdf wasn’t really designed to be modified in this way, if at all–at least according to my understanding. These problems would be alleviated if people switched to a different, more modular format, e.g. epub, and improved that rather than continue to use pdf in this way.
Edit: To add, pdfs are great for printing, but very annoying to work with on desktop–Windows especially–or edit.
Brilliant proposal to leverage existing capabilities and a mobile first mantra designers already know.
I have a folder with 160 random pdfs with following count of producersI guess hitting up Adobe first with a feature request might be a starting point.
[(‘Adobe’, 39), (‘cairo’, 16), (‘Recoded’, 14), (‘ABBYY’, 10), (‘LibreOffice’, 8), (‘Skia/PDF’, 6), (‘GPL’, 6), (‘Microsoft®’, 6), (‘calibre’, 6), (‘Qt’, 5), (‘Mac’, 5), (‘imagemagick. org’, 4), (‘OpenOffice .org’, 4), (‘Producer,’, 4), (‘iText’, 2), (‘iText®’, 2), (‘xdvipdfmx’, 2), (‘pdfTeX-1.40.17’, 2), (‘iOS’, 2), (‘iPhone’, 1), (‘PDFlib’, 1), (‘PDFsharp’, 1), (‘pdfTeX-0.13d’, 1), (‘CVISION’, 1), (‘BnF’, 1), (‘Prince’, 1), (‘pdfTeX-1.40.19’, 1), (‘Google’, 1), (‘freepdfconvert. com’, 1), (‘PDFium’, 1), (‘mPDF’, 1), (‘Some’, 1), (‘AFPL’, 1), (‘KONICA’, 1), (‘QuarkXPress®’, 1), (‘Nitro’, 1), (‘Python’, 1)]