[danso]

It depends on the backend workflow, no? If your system pulls in images and does the resizing, including the determination of horizontal vs. vertical, then outputting defined css selectors is straightforward.

However, if you're only in control of the front-end and are just reading from an API of photos where you're given a stream of image urls...how do you determine which css-selector (i.e. '.vert-image' and 'horz-image' with defined h/w dimensions) to put out? You can obviously just restrict everything to a one-size box, but either the vertical or the horizontal images are going to be given short shrift.