I've wondered for a while now why many books I have seem to use em units to specify the page margins instead of pixel units.
This is very annoying with professionally produced books from traditional publishing houses. To me, it looks like they try to reproduce the look of a regular paperback novel on the screen of the device (a PRS-505 in my case). This means significant amount of whitespace around the border of the page and fairly small text.
Unfortunately, the pixel density of e-ink screens isn't all that high yet, making text that small only marginally readable. So in many lighting conditions I have to magnify it to one or even two sizes larger. Since the margins are proportional to font height that means the originally large margins become even wider and at the largest font size I only get three or four words on a line, even though the text isn't very large!
Wouldn't it be better to use a fixed number of pixels for the page margins? Then they margins might be wider than I like, but at least they wouldn't grow even more
when the text is magnified.