|  | Posted by Jukka K. Korpela on 12/26/07 16:51 
Scripsit charles cashion:
 > I think I know what is happening.
 
 Part of it, yes.
 
 You seem to have formulated your question poorly, got flamed, and
 reacted by longish babbling. You have wasted everyone's time and, what's
 infinitely more important, my time. But hey, this is Christmas time, so
 we're are supposed to act as if we simulated attempts at being nice even
 to people who don't deserve it, so let's see... (To be honest, your
 question is interesting and relevant, and you didn't ask it in a
 completely obscure way.)
 
 > If the firefox browser does not recognize
 > face=roman or face=times or any other non
 > recognizable face, it defaults to  Times
 > New Roman.
 
 Not quite. It uses the browser's default font, which is, in fact, often
 Times New Roman, for some weird reason.
 
 This is close to a bug, but it can be explained as a feature. We would
 expect that with
 
 <font face="foo">bla bla <font face="bar">xxx</font></font>
 
 the text xxx would appear in the foo font, if the bar font is not
 available. If <font face="..."> refers to a font that is not available
 in the system, the markup shall be ignored (it is a suggestion that
 cannot be fulfilled), so the parent element's font shall be used.
 
 However, to justify the odd Firefox behavior, we can start from the fact
 that mapping of font names to actual fonts is browser-dependent anyway.
 Even on one system. different browsers may treat font names differently.
 For example, font names as such are case-sensitive, but browsers tend to
 treat them as case-insensitive, mapping e.g. arial, ARIAL and AriaL all
 to Arial. Moreover, they might interpret some names as variants of other
 names. Finally, a browser could map all unrecognized font names to a
 single font, namely the browser's default font; I cannot find anything
 that prevents this, even though it is odd and something that I wouldn't
 expect intuitively.
 
 So this more or less explains what I see on Firefox:
 When the font is set to Times New Roman, I see Times New Roman.
 When it is set to New Times Roman, Times Roman, or Roman, I see the
 default font I've set in Firefox (this time Verdana, just for testing).
 When it is set to Times, I see Times New Roman, which is very close to a
 serious bug, since my computer actually has Times, too, with slightly
 different characteristics (e.g., a wider "h"). But even this can be
 explained as browser-dependence: the interpretation of font names is not
 fixed. In practice, Firefox probably applies some "loose" font name
 matching, so that Times matches Times New Roman (even when Times
 actually exists on the system!).
 
 Internet Explorer 7 behaves in a somewhat more expected way:
 When the font is set to New Times Roman or Times Roman, I see Comic Sans
 MS, i.e. the font of the enclosing element. This is expected behavior:
 the unrecognized font name is ignored, in the sense that <font
 face="...">...</font> markup has no effect.
 When it is set to Roman, I see some rather ugly font, since my computer
 seems to contain such a font under the name Roman, labelled as "OEM/DOS"
 font. I guess it suffers from font smoothing, since it was created
 before modern smoothing techniques.
 And when the font is set to Times, I see Times - slightly larger than
 Times New Roman, with some differences in glyph design. Most computers
 probably lack such a font, and I don't really remember how I got it.
 
 What this means in practice is yet another reason why font suggestions
 may fail to have any effect or, worse, may have some unintended effect.
 
 > Now my question is: Is it possible to change
 > the *default* font?
 
 Of course. The way to do this depends on the browser and its version.
 It's typically something you do in a "Settings" window opened via a
 "Tools" menu. And you don't do it to other people's browsers unless they
 personally ask you to do that.
 
 > OBTW rf is correct. One cannot nest fonts.
 
 Who's "rf"? I'd hate to dig into my killfile. And if he or she has
 claimed that <font> elements cannot be nested, as I guess, he or she is
 completely wrong, once again, in a usual boring manner.
 
 --
 Jukka K. Korpela ("Yucca")
 http://www.cs.tut.fi/~jkorpela/
  Navigation: [Reply to this message] |