297076 - Firefox fails CSS1 test of absolute length units with regards to font-size

Status: RESOLVED INVALID

(Firefox :: General, defect)

Description

[Jackie]

Using: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b2) Gecko/20050526
Firefox/1.0+

According to CSS1 specifications (http://www.w3.org/TR/CSS1#length-units),
browsers should support absolute length units such as inches and centimeters.

Please note that for the following tests, your computer must be set so its
internal ruler reflects 'real-world'lengths, as described in the prologue to
these tests, at http://www.w3.org/Style/CSS/Test/CSS1/current/sec01.htm, in the
section entitled "Units and Monitor Testing."

All tests explain what behavior is expected according to CSS1 specifications.

Firefox passes tests involving absolute length units with regards to
text-indent, @ http://www.w3.org/Style/CSS/Test/CSS1/current/sec547.htm, and
margin(bottom and top)and padding (bottom and top), @
http://www.w3.org/Style/CSS/Test/CSS1/current/sec411.htm. 

However, Firefox fails when it comes to font size, as demonstrated in the
font-size test, http://www.w3.org/Style/CSS/Test/CSS1/current/sec526.htm. The
one cm and ten mm font-size was actually about 7 mm, the half-inch was about 3/8
in., and points was about 3/16 in. instead of 1/4 in. (one point = 1/72 inches).

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

What were you measuring?  The font-size is a measure of the em-square of the
font, which is whatever the font designer wants it to be.

Comment 2

[Brad Lassey [:blassey] (use needinfo?)]

The test is:
This sentence should be ten millimeters tall.

the paragraph tag for that specifies class="c"

and class c is defined as:
.c {font-size: 10mm;}

however, it displays  as 7mm

The same is true of the 1 inch and 1 cm tests.

(In reply to comment #1)
> What were you measuring?  The font-size is a measure of the em-square of the
> font, which is whatever the font designer wants it to be.

Comment 3

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

What are you measuring?  The height of a capital letter?  The distance from the
top of an É to the bottom of the descender of a p?

Comment 4

[Brad Lassey [:blassey] (use needinfo?)]

On firefox, running on Mac OS X 1.4, I'm highlighting the text and measuring the highlighted region.  It 
measures at ~8.5 mm


(In reply to comment #3)
> What are you measuring?  The height of a capital letter?  The distance from the
> top of an É to the bottom of the descender of a p?

Comment 5

[Jackie]

(In reply to comment #4)
> On firefox, running on Mac OS X 1.4, I'm highlighting the text and measuring
the highlighted region.  It 
> measures at ~8.5 mm
> 

On firefox, WindowsXP, highlighting the text and measuring the highlighted
region gives me .6 in. for the half inch, and 12 mm for the 10mm. 


> (In reply to comment #3)
> > What are you measuring?  The height of a capital letter?  The distance from the
> > top of an É to the bottom of the descender of a p?

I was originally measuring the height of a capital letter. 

The test (http://www.w3.org/Style/CSS/Test/CSS1/current/sec526.htm) does not
have any descenders in the sentences to be tested. 

Comment 6

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

The Windows behavior sounds right for typical fonts; the Mac behavior sounds
odd. Is your logical resolution configured correctly, or do you have the same
error ratio on the other tests?

Comment 7

[Jackie]

Measuring from the bottom of the highlighted region to the top of a capital
letter gives .5 in., which is correct. The rest of the measurements (cm, mm, and
points) are also correct if measured in this fashion.

Comment 8

[Brad Lassey [:blassey] (use needinfo?)]

On the Mac, measuring from the bottom of the highlight, to the top of the 'T' I get ~7/16 inch for the 1/2 
inch, ~.8cm for the 1 cm and ~8mm for the 10mm.  Mac has no configuration for "real dimentions" as it is 
supposed to ship configured correctly.

(In reply to comment #7)
> Measuring from the bottom of the highlighted region to the top of a capital
> letter gives .5 in., which is correct. The rest of the measurements (cm, mm, and
> points) are also correct if measured in this fashion.

Comment 9

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

(In reply to comment #8)
> Mac has no configuration for "real dimentions" as it is 
> supposed to ship configured correctly.

This is completely false.  Mac always assumes 72dpi.  We ignore that and assume
96dpi instead, by default.

Comment 10

[Brad Lassey [:blassey] (use needinfo?)]

(In reply to comment #9)
> (In reply to comment #8)
> > Mac has no configuration for "real dimentions" as it is 
> > supposed to ship configured correctly.
> 
> This is completely false.  

So where is the configuration?

quoting from the w3c test suite configuration notes (http://www.w3.org/Style/CSS/Test/CSS1/current/
sec01.htm):

"Under WindowsNT and the MacOS, the system should have already determined real-world distances for 
your particular monitor and its settings. It is assumed that the user agent will take advantage of this"

>Mac always assumes 72dpi.  We ignore that and assume
> 96dpi instead, by default.

Well...it would appear that this assumption would be incorrect

"you know what happens when you assume, you make an a$$ out of u and me"