- From: Brad Kemper <brad.kemper@gmail.com>
- Date: Thu, 21 Jan 2010 06:33:56 -0800
- To: Giuseppe Bilotta <giuseppe.bilotta@gmail.com>
- Cc: David Singer <singer@apple.com>, "Tab Atkins Jr." <jackalmage@gmail.com>, "robert@ocallahan.org" <robert@ocallahan.org>, www-style <www-style@w3.org>
- Message-Id: <C94087FA-4CEE-43D5-9248-EA354554AED8@gmail.com>
On Jan 20, 2010, at 9:19 PM, Giuseppe Bilotta wrote: > On Wed, Jan 20, 2010 at 11:10 PM, Brad Kemper <brad.kemper@gmail.com> wrote: >> Au contraire. I said "in cases where the content is made especially for the >> output device", as opposed to general Web content. So the author would know >> the monitor dpi and UA default zoom choice. > > Assuming 'device' here refers to _both_ the physical display _and_ the > UA, because otherwise there are no guarantees that the default zoom > chosen by one UA is the same as the one chosen by another UA. Sure. But the UA could be a standard Webkit browser running on a laptop or mobile phone, while the "monitor" could be a more custom rig built into eyeglasses or being projected as Super-HDTV onto IMAX screens. >> And in the case of user style sheets, these would be used where the user did >> know that the page scaling was always wrong when viewing Web pages in his >> browser. > > Agreed for this case. But for the former, I still think that setting a > ratio between CSS and real world units would be more appropriate _and_ > more robust than setting a more abstract zoom level. My motivation for making it part of the zoom level here is that familiar UI control already exist in most Web browsers today for changing the page zoom, which does about the same thing. I would at least want that to work to undo inappropriate UA assumptions, even if a style sheet was not used. But with a user style sheet, it could be "set and forget", "one and done". Once it is a property for some style sheets, then it could be put into any style sheet (with caveats). If it is a property that you can apply to the root element (which in effect, whatever it is called, acts similarly there to the IE 'zoom' property, but without the hasLayout business), then I don't see the benefit of restrictions to ONLY allow it on the root or BODY. That seems like extra work, to make it unlike other properties that work on any element, when it could be potentially useful on those other elements too. I guess if nobody but me wants it to be a general property, then maybe an @rule would be better? I think writing 'reference-length: 1px = 1px' is more confusing than 'zoom:1'. Which side is the CSS px and which side is the device pixel? It seems to do the same as 'zoom:1', but you are also assuming it would only be set on BODY or :root?
Attachments
- application/pkcs7-signature attachment: smime.p7s
Received on Thursday, 21 January 2010 14:34:37 UTC