The IBM page seems to have an ellipsis character in UTF-8, with bytes E2 80 A6. The web server is set to force all browsers to use the encoding iso-8859-1 regardless of what charset is stipulated in the html code. The browser uses the Win 1252 equivalents and displays …
To see what a web server is forcing, if anything, you can use http://web-sniffer.net/ On Jul 21, 2015, at 9:45 AM, Marcel Schneider wrote: > > I fear things have grown somewhat upside down, so I'll try to outline the > real scenario: > > 1 - I open the page, the horizontal ellipsis is displayed as … (of course I > don't know yet that it's a horizontal ellipsis...). > 2 - I remember my comment about the T-shirt and decide to check whether it's > accurate. Firefox shows me the page is in UTF-8 and that there is nothing > after "Our apologies". > 3 - After some trial and error, I save the page in Zotero and open the > folder. The only HTML file inside is declared as Windows-1252, and there is > the horizontal ellipsis. > 4 - I back up the original file, try modifying the charset value to utf-8 and > refresh the page, the … converts to a horizontal ellipsis. > > To answer your questions, I figure out that the page was written on a > Windows-1252 template but without sticking with this charset. U+2026 was > probably an autocorrect. So it was "produced using UTF-8" but "the webmaster" > must have published it under the old charset. > > The puzzling point is that Firefox tried UTF-8 and told me he's serious, but > "ate" the U+2026 while it used the native Windows-1252 to "display" it... > > I hope that some macro could enable the "webmasters" to rapidly update > websites, because resolving this "funny" "scenario" has cost me some "effort" > today! > > Marcel >
