234829 - (JapaneseEncoding) Firefox forgets about Japanese encoding in new-window instance.

Status: RESOLVED INVALID

(Firefox :: General, defect)

Description

[S Davies]

User-Agent:       
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.6) Gecko/20040206 Firefox/0.8

The encoding recognition in Firebird/fox is outstanding, I've found but there is
a little bug.  On a lot of Japanese sites with links which open in a new window
(new instance), Firefox will forget the encoding it's supposed to be using
(Japanese) and it'll use its 'default' encoding.  This will result in the page
being rendered in the weird ASCII-melange one would usually find only in
Internet Explorer when viewing Japanese sites.  The problem is solved by
manually selecting Japanese encoding via View > Character Coding > Japanese but
only until a new instance is opened again, where it forgets again.

I can't imagine why this would be expected behaviour as it is slightly
irritating, having to reset the encoding every time a new page on a Japanese
site is opened.  So I thought I'd report it as a bug just in case.

The conditions don't seem to be limited to my hardware, fonts or pretty much
anything else.  The only thing that might be called into question is the OS, XP.
 I can say this without doubt because both I and a friend (he on his own PC,
with a completely different setup, other than the OS) have seen this occur.  I
use the latest Firefox nightlies from the FTP site.  He's using 0.8 goldbuild
because he doesn't trust nightlies.  So this seems to be a global bug.

Nonetheless, bug or not, I think you deserve kudos for your encoding becuase
Firefox's Japanese handling is REALLY nice.

Reproducible: Always
Steps to Reproduce:
1. Go to any site with a co.jp URL (or possibly any site that uses Japanese
characters but this is the only way I've tested it).
2. Find a link that will only open in a new window/instance and not in the
current window.
3. The new window will be default ASCII/ISO encoding and not Japanese.  View >
Character Coding > Japanese fixes this until another new instance is opened,
which resets again.
Actual Results:  
Standard Japanese characters and text are replaced with a melange of strange
ASCII characters.  For example, a page of Japanese after this bug has reared its
head will transform into:

「獣華街」より
「第一位皇帝ポルビオス・チョーラ」
コミックマーケット59
無料配布ペーパー（人類獣人化計画12号）のカット
えーっと、このキャラのカラーは
95年に描いた同人誌の表紙以来…。
彩色はギャルゲー風に塗ってみました。
個性は無いけど見た目はイイかな？って。
実は髪の毛の部分をパーツ分けして彩色したので
PCにスゲェ負担掛かってヒヤヒヤものでした。 

Whereas before the new instance, it's perfectly alright and after setting
manually, it's perfectly alright.

Expected Results:  
This might merely be a feature that hasn't been implimented yet or a desired
action but it seems sensible that a new browser instance should carry the
encoding from the last browser instance it was called from.

If you're working on this then please disregard this bug report and sorry to
have bothered you with it.

Comment 1

[Robin Monks]

This bug has been dead for 4 1/2 months.  I'm changing it to invalid as the
build is now very old.  If this can be reproduced on a nightly or FireFox 0.9,
please restart this bug.

Thanks