[rfc-i] Scaling factors for graphics

[mellon at fugue.com (Ted Lemon)]

On Jul 12, 2014, at 7:52 PM, Joe Touch <touch at isi.edu> wrote:
> Consider the periodic table of typefaces:
> http://wallpaperswide.com/periodic_table_of_typefaces-wallpapers.html
> 
> The names will *never* be legible on a VGA display, regardless of whether it's 2" diagonal or projected on the side of a barn.
> 
> On a 4" iPhone display, the names are legible (barely) - granted they're small, but legible.

Ah.   This is only legible on my laptop display at full display resolution.   I would not be able to read it off a Retina display without a magnifying glass.   So I don't think it should be displayed on a Retina display shrunk to fit the screen even though in principle the information is there, which it wouldn't be on a vga screen of the same size.   That's what I was getting at.

You do have a point that pixel resolution matters, but I wasn't imagining that we'd even support down-scaling the image to fit on a low-resolution display.   So what I had in mind was that if the minimum width was specified as 4"x9", and the pixel resolution of the image was 1920x1080, then on a retina display it would still be scaled up and scrollable even if it fit, and on a 13" vga display, it would still be scrollable despite fitting the physical dimensions, rather than being down-scaled.

I'm sensitive to this issue because my Google phone has a 1920x1200 display, and stuff that fits perfectly well at pixel resolution is sometimes completely unintelligible because the image is so small.

However, given that we're talking about SVG (probably), specifying a minimum pixel size as well as a minimum physical dimension is a good idea.