Re: Bug in TR 19, and fancy HTML in TR's

From: Katsuhiko Momoi (
Date: Sat Jul 08 2000 - 20:27:59 EDT


I need to correct myself below. The problem remains a mystery.

- Kat

Katsuhiko Momoi wrote:
> Doug Ewell wrote:
> >
> > Michael \(michka\) Kaplan <> wrote:
> >
> > >> Can the authors (or reformatters) of Unicode Technical Reports please
> > >> keep the use of "fancy" HTML and Javascript in TR's to a minimum?
> > >> Tables and simple stylesheets ought to be sufficient. This is not an
> > >> elaborate sales or marketing presentation, after all -- it's a
> > >> TECHNICAL REPORT. Please don't tell me to update or change my browser
> > >> to get the "full experience" of the TR. I should be able to print the
> > >> thing on dead trees and get all the information that way if I choose.
> > >
> > > Well, some people might actually prefer you upgrade your browser to
> > > something that supports Unicode a bit more effectively, at the very
> > > least. :-)
> > >
> > > NN 4.0x has tons of problems with bidi and other complex scripts, with
> > > Asian characters, and more. Its not entirely unreasonable for the
> > > online resources of the Unicode Consortium to make use of technologies
> > > that support Unicode and not be obliged to test everything on older
> > > browsers that do not. I think its likely that they did not even know
> > > that such an issue would exist, since they have upgraded to browsers
> > > and technologies that support the very standard they are writing for.
> >
> > This has nothing to do with bidi, Asian characters, or any other aspect
> > of Unicode support. I know that some of the TR's, such as TR 21, "Case
> > Mapping," include UTF-8 characters and these generally do not print
> > correctly using older browsers. I know about this problem and deal
> > (reluctantly) with it. This is not about any Unicode-related problem,
> > but about advanced HTML. TR 19 is tagged as UTF-8, but contains only
> > one UTF-8 character (U+00A9 COPYRIGHT SIGN), and this is clearly not
> > where my problem lies.
> >
> > I understand the need and desirability to use Unicode features in a
> > Unicode technical report, I just don't see the major advantage of
> > embedding Javascript in one.
> >
> > I wonder if Erik or some other Netscape person can examine the TR and
> > figure out what NN 4.06 doesn't like about it.
> >
> The problem you cite in TR #22 (e.g. 2.2 History) and other parts which
> don't print well use the <PRE> tag. For example,
> <H3>2.2 <A name="History">History</A></H3>
> <PRE>
> <history supercedes="CP501" derivedFrom="CP500">
> <modified version="2" date="1999-09-25">
> Added Euro.
> </modified>
> <modified version="1" date="1997-01-01">
> Made out of whole cloth for illustration.
> </modified>
> </history>
> </PRE>
> Note that within the <PRE> tags, there are also other HTML-tag like
> elements. This is causing the problem -- NN is ignoring them in
> printing. (cf. Even within the <PRE> tags, you are supposed to regard
> tags like "<B>" as HTML elements and not show them.) I will put aside
> the question as to whether or not this is a bug in NN but from a
> practical point of view, the authors can avoid this problem by escaping
> the special characters, i.e. "&lth;" for "<" and "&gth;" for ">", within
> the <PRE> tags.

This comment of mine is not correct. When I looked at the source of the
document, NN normalized "&lth;" and "&gth;" into "<" and ">" in the view
source window, respectively. So the source is actually using the CERs
within the <PRE> tags. What makes this problem mysterious is that when I
extracted a section of this document, "2 XML Format", and created a new
document with just the copied section, all of the preformatted sections
within it printed correctly. Yet, when the section is embedded within
the rest of the document, these preformatted parts do not print out. I
looked through the rest of the document briefly but did not so far find
anything obvious that might cause this kind of problem. The results are
the same whether the document is printed from a locally saved document
or from the web directly. So for now, this remains a mystery.
BTW, Mozilla does print out the preformatted parts OK.

> > > As a side note, I do not test code on my Osborne 1 or my Epson PX-8,
> > > either. :-)
> >
> > Nor I on my Timex Sinclair 2000 or Atari 130XE. But I hardly think of
> > Windows 95 and NN 4.06 as being quite as ancient as that.
> >
> > -Doug Ewell
> > Fullerton, California

Katsuhiko Momoi
Netscape International Client Products Group

What is expressed here is my personal opinion and does not reflect official Netscape views.

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:05 EDT