Possible FireFox font-size bug: does not honor CSS font-size for code tag wrapped in pre

Possible FireFox font-size bug: does not honor CSS font-size for code tag wrapped in pre

I've found an inconsistency among browsers in handling font-size (on Mac OS X 10.11.4). I'd like to know if this is a bug in FireFox, or a bug in CSS, or am I not understanding something about CSS?

The jsfiddle Possible FireFox font-size bug shows the details. In a section like this:

Start of pre section    **problem here!**  End of pre section  
  

the style code { font-family: Courier; } changes the font-size shown by Safari and Chrome, but not in FireFox. Yet in other sections, the code element size increases from 13px to 16px in all the browsers.

How it looks in Safari (and Chrome) without any style:

How it looks in Safari (and Chrome) with code { font-family: Courier; }

How it looks in FireFox without any style:

How it looks in FireFox with code { font-family: Courier; }

Why does the font-size increase from 13px to 16px after setting the font-family in all browsers?

Perhaps FireFox is changing the font-family but not the font-size. Yet it does change the font-size in other places, like in code inside a p element or inside a list.

I'm guessing the solution is to just aggressively set explicit font size with pixel units wherever I care about it (ignoring accessibility).

---

Quote from possibly related FireFox bug from 2006: Bug 328621 - strange default `monospace' font size -- differs from the proprtional fonts

Though the inter-browser portability problem this imposes doesn't seem that trivial to me. The last time I wanted to markup HTML so that the monospaced parts showed the same in both Mozilla and M$IE I had to resort to absolute font sizes, which should be certainly avoided due to the accessibility problems they impose.

And yes, there's a simple workaround -- changing the settings. But most of the users will use defaults. :-(

---

The answer from @BoltClock was super helpful. The page he/she referenced about Fixed Monospace Sizing by Eric Meyer has a thorough analysis and some solutions. It explains some of the browser mechanics related to when the user's font preferences are used, which explains some of the mysterious behavior I was seeing (e.g. why was the code element seeming to inherit a size of 13px when it was the child of a p element with size of 16px: the reason is it actually inherits size medium which causes the browser to use the default browser/user preferences).

For my situation I'm finding I can remove the html { font-size: 16px; } style and instead add the following:

pre { font-family: monospace, monospace; }  code { font-family: monospace, monospace; font-size: 100%; }  

This gives me consistent relative font-sizing but still allows the user to set the default font-size. Eric Meyer recommends using monospace, serif but I'm finding that monospace, monospace works, and I think it makes a little more sense when you see it in a style-sheet (it's odd, but looks probably intentional, and it means you really, really want monospace).

I think that the ultimate problem is that the browsers have different font-size defaults/preferences for monospace font, than for other fonts. The main idea of "accessible responsive design" in regards to font-size is to let the user have control over font-size. But when we have several settings for font size (serif, sans-serif, monospace) then that's too many options for the user to set. The user just wants bigger text that they can read. But for the author: they just want the relative sizes of the fonts to be correct at whatever given size.

In case it's helpful to someone else, here are some examples (in the slick Eric Meyer style) showing that monospace, monospace fixes some of the problems I wrote about in my own answer below.

code:before {content: attr(style); border-bottom: 1px solid #AAA; color: #575; margin-right: 0.5em;}  p {margin-bottom: 2em;}

This is a 'p' element with a 'code' element contained within.
  
This is a 'p' element with a 'code' element contained within.
  
This is a 'p' element with a 'code' element contained within.
  
This is a 'p' element with a 'code' element contained within.
    
This is a 'p' element with a 'code' element contained within.
    
This is a 'p' element with a 'code' element contained within.
  
This is a 'p' element with a 'code' element contained within.
    
This is a 'p' element with a 'code' element contained within.

Answer by Johannes for Possible FireFox font-size bug: does not honor CSS font-size for code tag wrapped in pre

Why does the font-size increase from 13px to 16px after setting the font-family in all browsers?

probably because the browser's default stylesheet has these settings.

I'm guessing the solution is to not try to understand all this but just aggressively set font and font-size wherever I care about it.

Not sure what you mean by "agressivly" but yes: If you have particular demand for that style, set it as precisely as possible in your own stylesheet, including font familiy, weight, size etc.

Answer by Sparky256 for Possible FireFox font-size bug: does not honor CSS font-size for code tag wrapped in pre

I run into this problem often and it is a pain. Your application/browser can supply you with 'application' fonts, or 'system' fonts.

Your browsers are going by their default font type and size. An OS like Windows or Linux uses 1em, or 16px font size by default.

In my webpages I always set font-size to 62.5%, which makes 1em equal to 10px, 1.6em equal to 16 px, etc. I suggest you look at the size settings for the browsers, any web pages you created, and the current font size of the OS.

I have my OS set to 125% so I can see small icons and text. It does NOT affect webpages where I have defined the font size, but it does affect MSword and Notepad and other system apps. Understand they are separate per PC and browser, and some apps.

Just like I do, you may have to adjust for each instance of use.

Answer by owler for Possible FireFox font-size bug: does not honor CSS font-size for code tag wrapped in pre

Since I asked the question and got a couple of good faith answers that are however not really answering my question: I'm going to try to prove that this really is a problem with current browsers. See my comment at the end about why this might or might not matter.

Here then are

3 problems with font-size in CSS/HTML in FireFox, Chrome and Safari on Mac OS X 10.11.4:

Problem 1. Changing font-family should not change font-size, but it sometimes does.

I don't know what the "user agent stylesheet" is. There can be all kinds of settings on it. But regardless of those settings, changing font-family should not change computed font-size. (Some fonts might appear larger at the same font size compared to other fonts at same font size, so visually they can look bigger or smaller, but the computed font size number should not change).

notation: in the following

pre(code) means roughly 
text here
  p(code) means roughly 
text here
  

In the default situation, a page with no styles, I see this in all 3 browsers:

pre=13px  pre(code)=13px  p=16px  p(code)=13px  

pre pre(code)
  
paragraph p(code)

Perhaps the user agent stylesheet is something like this:

pre { font-size: 13px }  code { font-size: 13px }  p { font-size: 16px }  

The user agent can't be using relative sizes for code, because it appears as 13px inside of both pre and p elements. So it could be this:

pre { font-size: 81.25% }  code { font-size: 13px }  p { font-size: 100% }  

or the equivalent using em instead of percentage. Or perhaps a descendent rule with percentages or em:

pre { font-size: 81.25% }  pre code { font-size: 100% }  p { font-size: 100% }  p code { font-size: 81.25% }  

But then how do we explain this:

After setting code { font-family: Courier } (or any font family other than monospace) we get:

pre=13px  pre(code)=16px for Chrome & Safari    pre(code)=13px for FireFox  p=16px  p(code)=16px  

code { font-family: "Courier"; }

pre pre(code)
  
paragraph p(code)

No matter what the stylesheet is, the font-size should not change after only changing the font-family.

Problem 2. Relative font-size should be based on parent element's computed font-size. But it sometimes is not.

With no styles we have:

p=16px  p(code)=13px  

After setting code { font-size: 100%; } we should have

p=16px  p(code)=16px  

Instead we still have:

p=16px  p(code)=13px  

This is wrong because author style should override default user-agent (browser) style. And relative (percentage) font-size is based on the parent element's computed font-size.

code { font-size: 100%; }

paragraph p(code)

Similarly, after setting code { font-size: 120%; } we should have

p=16px  p(code)=19.2px = 16px * 1.2  

Instead we have:

p=16px  p(code)=15.6px = 13px * 1.2  

code { font-size: 120%; }

paragraph p(code)

This shows that parent element is being ignored, and that some phantom value of 13px is being used to calculate the relative font-size.

Problem 3. The standard advice to use html { font-size: 62.5% } and then use em or rem for font-sizing does not work correctly

(Note: FireFox with rem does seem to be correct)

(Note: using the also recommended body { font-size: 62.5% } seemed to produce some crazy results, like doubling of font sizes, so I'm not considering that here.)

With html { font-size: 62.5%; } and code { font-size: 1.6em; } we have:

p=10px  FireFox: p(code)=13.0167px  Chrome: p(code)=13px  Safari: p(code)=13px  

These should all be 16px = 1.6 * 10px

body { font-size: 62.5%; }   code { font-size: 1.6em; }

paragraph p(code)

With html { font-size: 62.5%; } and code { font-size: 1.6rem; } we have:

p=10px  FireFox: p(code)=16px  Chrome: p(code)=13px  Safari: p(code)=13px  

FireFox is the only browser to get it right here.

html { font-size: 62.5%; }   code { font-size: 1.6rem; }

paragraph p(code)

I suspect this all doesn't matter much because people are used to slapping font-size styles all over the place in their CSS.

From my reading, there has been an attempt to make web pages more responsive to the native settings on each device: to adapt to varying hardware with more or fewer pixels per inch (PPI); to allow users with reduced vision to request larger text so they can read more easily; etc.

There are various recommendations and a bit of controversy about how to best do this: specify font-size with html { font-size: 62.5% } and use em or rem everywhere seems to be a popular opinion. Lately just using good old px to specify font-size is coming back in style, with the idea that px does not specify actual screen pixels but rather an idealized 96 dpi for the viewer (which may translate to a different number of pixels) and letting the browser's zoom controls handle user desired changes in font-size.

In summary, I think these problems with font size argue against the logic behind most of the advice about "responsive design" for font size. My guess is that this is an area where there is a lot of history factored into how browsers work, and it's near impossible to change now. And of course there are millions of websites with existing code that are all presumably happy with how font sizing is working now.

It seems to me that the only way to get consistent font sizing is to use pixels: px. For example, px at the Root, rem for Components, em for Text Elements. Or just just use pixels everywhere.

Mozilla writes the following about accessibility and font-size

Defining font sizes in pixel is not accessible, because the user cannot change the font size from the browser. (For example, users with limited vision may wish to set the font size much larger than the size chosen by a web designer.) Therefore, avoid using pixels for font sizes if you wish to create an inclusive design.

But with all these bugs with inherited and relative font size it doesn't seem like an option to avoid using pixels. Plus the argument about a reference pixel and using the browser's zoom control seems to solve the accessibility issue.

Answer by BoltClock for Possible FireFox font-size bug: does not honor CSS font-size for code tag wrapped in pre

This issue is not new; it has been known for many years that browser handling of monospace fonts is incredibly screwy and inconsistent across browsers. A number of workarounds are available that don't require you to override the monospace font preference set by the user, the most perplexing of which is to specify

font-family: monospace, monospace;  

That's right: two monospace keywords at the end of the font stack. Nobody knows why this works; it just does.

For what it is worth, Firefox's UA stylesheet contains references to a -moz-fixed keyword which refers to the preference that is set by the user, which includes both the family and the size despite it being a value for the font-family property. Firefox ships with the preference set to 13px in whichever the system's default monospace font is. It would seem that monospace, monospace forces the browser to compute the element's font-size according to the spec while still preserving the preferred monospace family, at least. But this is just a guess.

Is this a bug? It depends on whom you ask. If you ask browser vendors, they'll probably say this is intentional. If you ask other authors, they'll probably also call it a bug. Is this a spec violation? No, because the spec allows browsers to implement defaults however they like.

Answer by nim for Possible FireFox font-size bug: does not honor CSS font-size for code tag wrapped in pre

That's because back in the dark ages one Mozilla dev decided to enforce his personal preferences in the browser settings (most coders are young people who think it is smart to destroy their eyes by setting text size one or two sizes smaller than everyone else).

Thus, Firefox ships with defaults that alias "monospace" to a smaller font size that "sans serif". Pretty much destroys coordinated font sets (where the monospace font is designed to look the same as the serif or sans serif one).

You can "fix" it locally by launching about:config and changing all font.size.fixed.* properties to the same value as font.size.variable.* ones. But since you can't fix all Firefoxes in the wild, that makes monospace useless for this browser.

Fatal error: Call to a member function getElementsByTagName() on a non-object in D:\XAMPP INSTALLASTION\xampp\htdocs\endunpratama9i\www-stackoverflow-info-proses.php on line 72