status o' the day

Thomas Dowling tdowling at OHIOLINK.edu
Sun Feb 9 11:37:44 EST 1997


Karen Schneider writes:

> I have to say that one thing I do knowing it is wrong is embed tags
> inappropriately.  But may I humbly ask the master, if "<h1><b>Really</b>
> Important Header</h1> doesn't work," does that mean you know it breaks
all
> browsers or that it just ain't right, HTML-wise?  

I vaguely recall seeing something like this break one browser.  I'm trying
to remember where I saw it, and it's probably an obsolete version of some
weird little browser (a voice in the back of my head says Oracle
PowerBrowser, but I've been trying to ignore the voices in my head), but I
know I've seen a browser that does this:

  <h1> (Browser says "Oh, this means turn on bold text in some
appropriately large font")
    <b> (Browser says "Oh, this means turn on bold, which is already on, so
do nothing")
      Really  (Browser renders this in bold)
    </b> (Browser says "Oh, time to turn off bold text")
      Important Header (Browser renders this in large font, but not in bold
which just got turned off)
  </h1> (Browser reverts to normal sized font)

Since most browsers already render headings in bold, I don't know what the
author of the above would expect to accomplish, but it's taken from a real
example.  Nesting <I> inside a header might make more sense, but it's not
inconceivable that someone would configure a browser to render second- or
third- level headers in italic text (it's the default style in Word as I
recall, and might well turn into some kind of Word Internet Assistant
stylesheet in the near future, as one example).

As I mentioned before, violating any DTD isn't exactly a crime.  It
sacrifices a degree of predictability over how pages are rendered for a
percentage of users.  If you can live with both those figures, go ahead. 
But you should do so knowledgeably rather than letting your copy of Hot
Front Gold Page Pro do it for you.

Thomas Dowling
Ohio Library and Information Network
tdowling at ohiolink.edu



More information about the Web4lib mailing list