244330 - Switching to Unicode UTF-16 causes Firefox to display garbage

Status: RESOLVED INVALID

(Firefox :: General, defect)

Description

[Eric Murphy]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a) Gecko/20040515 Firefox/0.8.0+
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a) Gecko/20040515 Firefox/0.8.0+

See screenshot.

Reproducible: Sometimes
Steps to Reproduce:
1. Go to http://slashdot.jp
2. View -> Character Encoding -> Unicode

Actual Results:  
See the screenshot. Whole page is rediplayed as garbage. Switching to a
different character set will show the page again.

Expected Results:  
Try to display some sort of representation of the page other than a page full of
garbage characters.

I do not have the Japanese fonts to display the page, could this be a trigger
for this problem? The page is served up as Japanese character set.

Comment 1

[Eric Murphy]

Attachment: Actual Result

Comment 2

[PikeUK]

At the risk of revealing my nearly total ignorance of this subject, surely if
the page is encoded in EUC-JP and you ask Firefox to intepret it as UTF-8 then
garbage is what should be displayed, or am I missing something?

Either way this bug should probably be in the Browser product.

Comment 3

[Eric Murphy]

Yes, that is true, but just for the characters. However in this case, the whole
page is lost. 

Comment 4

[L. James Sprague]

Whenever I get that full page of garbage, I notice it's because the encoding
became set to Unicode 16, and switching back to 8 reveals the page. I don't know
how it ends up on Unicode 16.

I additionally have a problem that is certainly related. When I File > New Tab I
get, instead of blank page, a line of garbage. The encoding is inexplicably UTF
16 again, and switching that to 8 makes the space blank again. 

This sometimes even happens right before a page loads. For instance my homepage
is set to google.com and when I start Firefox up, a line of garbage is displayed
for a half second before Google appears.

Comment 5

[Eric Murphy]

I was not specifying that this problem is UTF-16 only. Thanks Comment #4.

Comment 6

[Eric Murphy]

I don't know what I was smoking before or if the problem is worse now, but this
problem seems to happen on every web page all of the time. Removing URL.

Reproducible: Always
Steps to Reproduce:
1. Go to a web page.
2. View -> Character Encoding -> Unicode (UTF-16 Little Endian)

Also see http://bugzilla.mozilla.org/show_bug.cgi?id=171984 as that problem may
be related.

Comment 7

[Jungshik Shin]

> Switching to Unicode UTF-16 causes Firefox to display garbage

This bug as filed is invalid. If you manually switch to UTF-16 when the actual
encoding used is not, what do you expect other than the garbage? 

If it's about UTF-16 being automatically chosen when it should not, please
change the summary line accordingly. Otherwise, this bug has to be closed as
invalid.

BTW, what's your default character encoding? (Is it set to UTF-16?) Did you turn
on the automatic detector? Does it happen with a recent nightly build of
Mozilla, too? 
 
BTW, please do not use the full URL when refering to another bug (bugzilla
doesn't show the tooltip if a bug is cited with the full url). Just use the bug
number preceeded by 'bug' like this: bug 171984

Comment 8

[Simon Montagu :smontagu]

Comment 2 is correct but needs some filling out. If you try to display a website
in the wrong encoding, it appears as garbage to a lesser or greater extent,
depending on how much the two encodings overlap. 

In the example of EUC-JP and UTF-8, as comment 3 says, all the japanese
characters will appear as garbage but the structure of the page and any words in
Latin script will still appear correctly, since the US-ASCII range is the same
in both encodings.

In the example of EUC-JP and UTF-16 there is no overlap at all, so you get the
full page of garbage shown in the attachment.

This bug is INVALID. Comment 4 seems to be about something different, maybe an
autodetection issue.

Update: jshin and I just midaired saying much the same thing. I think this bug
should be closed anyway, and if there is a genuine bug with UTF-16 being
selected when it shouldn't be (per comment 4) it should be filed separately.

Comment 9

[Eric Murphy]

Ok, that sounds like a good explanation of the reason why this is happening.
However, should the application somehow notify the user the reason why the WHOLE
page (layout) is garbage, rather than just the text of the page.

When it happens it looks like a bad bug in the browser and the user may think so
too (as I did).

I have not had the UTF-16 encoding switch on by itself. Apparently the other
person has. Yes that should be a separate bug.

Comment 10

[JongAm Park]

Even when the page is written in English, just like the Mozilla bug report page,
if you choose the UTF-16, it shows garbage.

Confirmed with Mac version of the Firefox.

Comment 11

[Jungshik Shin]

(In reply to comment #10)
> Even when the page is written in English, just like the Mozilla bug report page,
> if you choose the UTF-16, it shows garbage.

English or not, it doesn't matter.