Re: cp1252 decoder implementation

May be you've forgotten FrontPage, a product acquired by Microsoft and then
developped by Microsoft and widely promoted as part of Office, that
insisted in declaring webpages as "ISO 8859-1" even if they contained
characters that are only in "windows-1252". Even if we edited the page
externally to correct it, FrontPage *automatically* changed it into
"iso-8859-1".

So Microsoft has a responsability about this, even if the problem first
appeared in Netscape browser (but at least Netscape accepted pages declared
as "iso-8859-1" without problems, as well as web pages correctly labelled
as "windows-1252".

I perfectly recall the many support questions sent to MSDN about this
Frontpage bug (and the many complains received by webdesigners using
FrontPage, that their websites were bogous, and that these webdesigners
were "stupid", "incompetent", never "read the RFC's" and so on...)

Microsoft never published a correction in FrontPage, pretending that this
would cause problems of interoperability (but interoperability problems
were already existing at this time : only IE could then correctly display
pages edited in FrontPage, Netscape at this time was only working in
Windows, and on all other OS'es those pages were garbled)

I do think that this is this FrontPage bug that created the proliferation
of webpages incorrectly declaring the wrong charset, and that then forced
all browsers to adapt. The Netscape bug itself did not initiate this
proliferation (and Netscape was even corrected later without this bug, but
then had to revert this correction due to the proliferation of incorrectly
declared webpages).
Received on Wed Nov 21 2012 - 14:25:24 CST