W3C home > Mailing lists > Public > www-international@w3.org > October to December 1999

Re: Fw: HTML versions and Browser compatibility

From: Nir Dagan <nir@nirdagan.com>
Date: Tue, 09 Nov 1999 16:56:10 -0500
Message-Id: <199911092154.QAA19226@vega.brown.edu>
To: "Suzanne Topping" <stopping@rochester.rr.com>, "www" <www-international@w3.org>
At 10:53 AM 11/9/99 -0500, Suzanne Topping wrote:
>> Apologies if this is simplistic question, but I am wondering if a
>> compatibility
>> list exists somewhere which shows which HTML versions work "properly"
>> with which browsers and versions? (Properly in this context meaning with
>> the fewest bugs.)

A simplistic answer is that HTML2.0 and HTML3.2 work with almost all browsers.
And that the following features of HTML4.0 are un-evenly supported:

1. i18n: transforming bytes into characters, understansing character entities,
   BiDi...
2. the <object> element. 
3. accessibility features: special attributes to tables and forms, among
other features.

What you'd realy need to localize a site is to have a very detailed list of
the
particular requirements of your "local language" that are in HTML4.0 are
actually 
supported in browsers. You almost have to go script by script, browser by
browser, 
to give a practical answer.

>> 
>> Also, (and I'm sure this is another simplistic question) if a site is
>> developed
>> using one version of HTML, is it possible to change that version during
>> localization to ensure wider character support?

The simplistic answer is yes.
To ensure wider character support in the HTML version, you have to change 
the SGML declaration to allow for more characters. This doesn't require any 
change in the SGML DTD of the HTML version in question. 

You may also define in the DTD more entities if you want those to refer to 
characters, rather than using numerical references.

This clearly doesn't ensure anything about what browsers will do with your
HTML.
It will only ensure that an SGML parser that uses your SGML declaration and
DTD 
will parse the document correctly.

If I were to write a site not in in the Latin script I'd use HTML4.0, but may 
trim it down to exclude <object> and possibly other poorly supported
features. 

Just one more thing, the character reference &#150; mentioned below is 
not defined in HTML and should not be used. As opposed to what Otto Stolz is 
implying, it cannot be used even if one is encoding his pages in 
"a proprietary codepage."  The "proprietary codepage" only defines how 
bytes are transformed into characters. And there is no 150th character in 
HTML. en-dash can be written in HTML as &ndash; or &#8211; or &#x2013;

Regards,
Nir.

>> 
>> Thanks for your tolerance and replies.
>> 
>> Suzanne Topping
>> Localization Unlimited
>> 
>> ----- Original Message -----
>> From: Otto Stolz <Otto.Stolz@uni-konstanz.de>
>> To: Unicode List <unicode@unicode.org>
>> Cc: Unicode List <unicode@unicode.org>
>> Sent: Tuesday, November 09, 1999 10:01 AM
>> Subject: Re: Non-breaking hyphens and web browsers
>> 
>> 
>> > Am 1999-11-06 um 19:41 h hat Ben Yenko-Martinka geschrieben:
>> > > While the en-dash "&#150;" [...] appears to work fine in Netscape,
>> > > it allows wrapping in Microsoft Internet Explorer.
>> >
>> > This is a character from a proprietary codepage, cf.
>> > <http://czyborra.com/charsets/codepages.html#CP1252>.
>> > Most probably, it will be understood only in MS-Windows systems --
>> > not in Unix boxes, not on Macs, probably not even on PCs running DOS,
>> > Linux, or OS/2 (I haven't tried these latter, though).
>> >
>> 
>> 
> 
===================================
Nir Dagan
Assistant Professor of Economics
Brown University 
Providence, RI
USA

http://www.nirdagan.com
mailto:nir@nirdagan.com
tel:+1-401-863-2145
Received on Tuesday, 9 November 1999 16:54:28 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 19:16:54 GMT