174855 - Autodetct charset does not work correctly with meta content tag

Status: VERIFIED WORKSFORME

(Core :: Internationalization, defect)

Description

[toto]

How to reproduce:
1. Set View->Character coding-> Autodetect to Russian.
2. goto url above
3. After the page loads, watch how View->Character coding is set by Mozilla to     "Cyrillic ISO-8859-5"
 (or if you know russian, watch that the page is rendered in incorrect encoding).
4. Right click on webpage -> View page source and find the tag 
   <meta http-equiv="content-type" content="text/html; charset=UTF-8">,
   so the page must be autodected as utf-8.
Reproducible: always.
Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826
on windows 2000 rus + sp3.
I think page was loading ok with mozilla 1.0, but i am not sure.

Comment 1

[Boris Zbarsky [:bzbarsky]]

With autodetect set to "Russian", the page detected as UTF-8 (I had to change
the "interface language" setting to "russian" in Google preferences to test this
properly).

This is with Linux trunk build 2002-10-14-08.

Comment 2

[Simon Montagu :smontagu]

Autodetector issue to shanjian.

Comment 3

[Yuying Long]

Works for me with any auto-detection on http://www.google.com has meta-tag
charset=utf-8.

Comment 4

[toto]

The source of bug is www-proxy-server (Proxomitron) thru which i must access www.
Works for me.

Comment 5

[toto]

Strange, but internet explorer works ok with or with out Proxomitron proxy.

Comment 6

[toto]

Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.2b) Gecko/20021030
Finally works for me. Closing the bug.

Comment 7

[Yuying Long]

Mark as verified per comment above.