RE: Flash of unstyled text/Flash of unfonted content

Wednesday, October 14, 2009 <>:

>My point was about web documents for which the fonts were essential,  
>rather than a style choice - 

I would argue that the fonts have always been essential, but I know what you
It's the change from fonts being OS centric to being server or "document"

If glyphs critical to an understanding of the page are missing - or, to look
at the more pessimistic flip side - glyphs are missing that might lead to a
*misunderstanding* of the page, then the page should fail in a very obvious
Currently, I believe, the UA will display question marks in place of missing
Is this sufficient? I don't know. I'd like to hear from some implementors on
They might be a few steps ahead...

>this will be a new experience for publishers.
Yes, and a singularly significant inflection point in the web's maturation
as a publishing medium equal to, and in most ways, superior to print. It
rates the time spent getting it right.



-----Original Message-----
From: [] On Behalf Of
Laurence Penney
Sent: Wednesday, October 14, 2009 1:54 PM
To: Dave Crossland
Cc: www-font
Subject: Re: Flash of unstyled text/Flash of unfonted content

My point was about web documents for which the fonts were essential,  
rather than a style choice - this will be a new experience for  
publishers. If a map required characters such as U+F421 (PUA) from a  
symbol font, then without that font you get lots of missing character  
glyphs instead. For fonts containing minority language characters, you  
get missing glyphs on text using those characters. JavaScript in the  
document itself is probably the way to verify loading of absolutely  
essential items, though.

- L

On 14 Oct 2009, at 18:04, Dave Crossland wrote:

> Hi,
> Given text ui browsers or other cut-down browsers for embedded  
> platforms - wikibook? :) - and other specialists, 'fail to load'  
> seems very odd to me. If the html document comes down the line, but  
> linked resources dont , that doesn't seem a loading failure to  
> me...? Happy to be wrong on this...
> Regards, Dave
> On 14 Oct 2009, 3:57 PM, "Laurence Penney" <> wrote:
> On 14 Oct 2009, at 15:02, Richard Fink wrote: > > There's a lot of  
> gray area. Plus, there is a clear...
> It's not just aesthetics of course. Sometimes a webfont may supply  
> characters that are not in the system fonts, and therefore required  
> for the document to make sense. This could be for language reasons,  
> or for sets of sorts such as map symbols.
> Do we therefore need a CSS rule that somehow declares that a  
> document should be marked "failed to load", if a vital font could  
> not be accessed? Are fonts inherently any different from images in  
> this respect?
> - L

Received on Wednesday, 14 October 2009 19:53:22 UTC