font-weight issue

[Siemova]

I'm running 4.0b3 under Windows 7, and have stumbled across a bug. After the latest update to Gmail, for some reason the "Archive" button now consists of a b tag nested inside a div that has the css property font-weight: bold. Ordinarily this redundancy wouldn't cause a problem, but in this case the text is being bolded twice. I played around in Firebug (on this and other pages) and discovered the following:

• The issue applies to both strong and b tags.
• Of the "web fonts," it only seems to affect Arial.
• Bolding doesn't stack if the style is on the element itself -- only if it's on an ancestor element.
• Doctype doesn't appear to matter.

According to several coworkers, this does not happen in 3.x.

[LIMPET235]

Moving to Builds...

[Bluefang]

Personally, I think that's the way it should work. What other reason would you want to put a bold style in the middle of another bold style.

If you think it's a bug, feel free to file one
https://bugzilla.mozilla.org

[patrickjdempsey]

Is there a font-size increase somewhere? Some fonts appear to go bold in larger sizes. Maybe I am missing something here, but I didn't think there was such a thing as a double-bold font.

[Bluefang]

There are varying levels of boldness (wight), though AFAIK it has to be supported by the font or synthetically computed
http://www.w3.org/TR/CSS21/fonts.html#font-boldness

So, to me, it makes sense that stacking weighting elements would relatively adjust the font weight.

[patrickjdempsey]

Yes, there are the different numerical weights, but "bold" is a specific font-weight, typically weight 700. If you want something relatively more bold, you should use "bolder". The same is true for the difference between font-size "large" and "larger", one is specific and one is relative:

http://www.w3.org/TR/CSS2/fonts.html#font-boldness
https://developer.mozilla.org/en/CSS/font-weight

Redundancy is pretty common in style sheets. That looks like a bug to me.

[Siemova]

Yes, I'm afraid it's clearly a font-weight issue (unrelated to font-size), out of keeping with normal style behavior, and would make the lives of web developers like myself more frustrating if left unfixed. I will indeed file a bug if nobody else has yet.

[supernova_00]

This is an issue with d2d/dw that won't be fixed. I tried very hard but the sites need to change their code to be correct.

https://bugzilla.mozilla.org/show_bug.cgi?id=548975
https://bugzilla.mozilla.org/show_bug.cgi?id=550128

Read both bugs from start to finish.

[patrickjdempsey]

From bug 548975:

This correct rendering, it's caused by the markup which is requesting a font
that's bolder than the bold face of Arial. The testcase is equivalent to the
markup below:

Code: Select all

<div class="style: font-family: Arial; font-weight: bold;">
  <b>Bug 548975 - This is very bold with Direct2D enabled</b>
</div>


The <b> tag is has 'font-weight: bolder'. ..

Aha, so it's caused by the B markup being equivalent to "bolder". Interestingly, that breaks HTML 4.1 which considers B to be "bold":

http://www.w3.org/TR/html401/present/gr ... l#h-15.2.1


HTML 5 may represent a change in that, (although this document does not specify it):

http://www.w3.org/TR/html5-diff/#changed-elements

[Siemova]

Interestingly, that breaks HTML 4.1 which considers B to be "bold"

Interesting indeed. The same assumption is even more apparent in the 3.2 reference spec:
http://www.w3.org/TR/REC-html32-19970114#font-style

In consideration of this -- not to mention the common understanding of <b> as shorthand for "bold" (which I know the W3C is trying to get away from with HTML5, but is nevertheless etymologically accurate) -- I wonder why Firefox devs decided to make it equivalent to "bolder" in the first place.

IE appears to default font-weight to "bold" for those elements, but I see that Chrome also uses "bolder." Hrm.

[patrickjdempsey]

It could be up to interpretation of the HTML 5 spec, that B should be used to *always* offset the specified text from the surrounding text, and not just to be shorthand for BOLD. So if you markup a section of a header with B, that section will stand out even if the entire header is already in BOLD. It makes sense, but it's definitely a redefinition of what B means.

[Bluefang]

If the behavior of elements goes against the W3C spec, you can probably build a better case for having the behavior changed. And I don't recall that argument being used in either of those bugs.