The "automatic zoom" seems to regularly zoom to a level that will cut off text far down on a page. I'm not sure how "automatic" it is as it mostly seems to be a slightly zoomed in level compared to page fit. From the name it sounds like it should detect where there is text on a page and hide some air in headers/footers. However, even in cases where it cuts off right as the text cuts off, it's often unclear if there's a following line or empty footer space following the last visible line. As it is I'll switch to page fit for default, but it would be preferable if the out-of-the-box experience either did a better job of not cutting off right before the text ends; at that point both page fit and width fit are better default options. The examples are from a [https://einnsyn.no/api/v2/fil?iri=http://dokumentobjekt.acos.no/70004525-c564-48d9-8591-2cf8d05df235 zoning proposal in Oslo].