This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
In Internet Explorer 10's implementation of the spec, the min-width and max-width are set using physical pixels. That means that, even if your display uses a different DPI setting, this will have no effect on the viewport setting. This may just be a bug in Microsoft's implementation of the spec. However, there doesn't seem to be a clarification in the spec on how different DPIs affect the @viewport's behaviour. IMO, IE10's implementation is wrong, as it is inconsistent with how CSS pixels work.
(In reply to comment #0) > In Internet Explorer 10's implementation of the spec, the min-width and > max-width are set using physical pixels. That means that, even if your > display uses a different DPI setting, this will have no effect on the > viewport setting. > > This may just be a bug in Microsoft's implementation of the spec. However, > there doesn't seem to be a clarification in the spec on how different DPIs > affect the @viewport's behaviour. The DPI might affects the reference (CSS) pixel as specified in CSS Values and Units (referenced by Section 2 of the Device Adaptation spec). Pixel units are CSS pixels in CSS specs unless otherwise specified. I don't think it needs further clarification. > IMO, IE10's implementation is wrong, as it is inconsistent with how CSS > pixels work. I don't have an IE10 available, but it sounds like it's incorrect. In short it should behave like this example: Given two devices: 1. Window width of 640 device pixels and a 2:1 relationship between device and CSS pixels. 2. Window width of 320 device pixels and a 1:1 relationship between device and CSS pixels. This rule should set the viewport to twice the width of the window for both devices (since both will have an initial viewport (window) width of 320 CSS pixels): @viewport { width: 640px }
If no further clarification is required in the spec, I'll mark this as WONTFIX.