W3C home > Mailing lists > Public > www-style@w3.org > January 2010

RE: Making pt a non-physical unit

From: Alex Mogilevsky <alexmog@microsoft.com>
Date: Mon, 11 Jan 2010 05:38:47 +0000
To: "L. David Baron" <dbaron@dbaron.org>
CC: "robert@ocallahan.org" <robert@ocallahan.org>, www-style <www-style@w3.org>
Message-ID: <CF67E24A6230D0418F669B711BD5AADB10E676E3@TK5EX14MBXC110.redmond.corp.microsoft.com>
Apparently the whole point of discussion is a concept of treatment of 'pt' that is different from other units when it relates to actual display resolution, and it is a kind of treatment that I was not aware of. I have been paying attention to this thread but I am still not sure what exactly is the relationship of 'pt', 'px' and 'in' are in Mozilla and what becomes problematic when dpi is different from 96.

In IE definition of units, 1pt is exactly 1/72 of an inch, as well as 1mm is 1/254 of an inch. 1px is a unit that used to be more complicated but I believe it is now 1/96 of an inch (in any mode or zoom). So for the purposes of this thread I think it means that all units are equally physical or non-physical, as their relative size never depends on device or zoom.

At some point in this thread there was a notion that CSS2.1 would have to change to agree with pt being always proportional to px. Does it?

-----Original Message-----
From: L. David Baron [mailto:dbaron@dbaron.org] 
Sent: Sunday, January 10, 2010 6:26 PM
To: Alex Mogilevsky
Cc: robert@ocallahan.org; www-style
Subject: Re: Making pt a non-physical unit

On Tuesday 2010-01-05 17:45 +0000, Alex Mogilevsky wrote:
> For the record, IE treats “pt” the way you propose for a number of 
> versions already. At this time there is no physical units in IE (that 
> is there is no unit that doesn’t grow or shrink with zoom) .

This discussion isn't about whether the units grow or shrink with zoom, though.  I think we all accept that zoom is something that happens after we've processed the meanings of length units, so that it zooms all lengths.

> New units have been proposed that would be defined as strictly 
> physical (I believe somebody has proposed “dpx” for “display pixel”).

Even the proposed physical units (e.g., all but pt and pc, if we go with that proposal) would respond to zoom, as I think should any unit for display pixels.

The issue here is what the meaning of a 'pt' is at a zoom of 1, and whether the number of display pixels in a 'pt' should depend on the physical size of the display (in cases other than where the size of the display causes the number of CSS pixels per device pixel to
change) or whether it should have a hard-coded relationship to CSS pixels.

(So the relevant question for IE behavior is whether it honors the settings in Windows preferences that set the dpi to be used for font sizes.)


L. David Baron                                 http://dbaron.org/

Mozilla Corporation                       http://www.mozilla.com/

Received on Monday, 11 January 2010 05:39:22 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:07:42 UTC