173448 - Pref "network.cookie.cookieBehavior" not set automatically in Phoenix

Status: VERIFIED DUPLICATE of bug 170705

(Firefox :: General, defect)

Description

[Matthias Siller]

Different sites, which work fine with Mozilla, do not reload parts of the site
properly. As this reload is done using JavaScript. You can verify this bug by
going to http://derstandard.at/ with Mozilla and Phoenix and see the difference.
I found this bug on Windows NT and Windows XP.

Comment 1

[Asa Dotzler [:asa]]

Matthias, you tried the latest trunk Mozilla and Phoenix nightly builds?  Always
include the build id of the builds you tested. Thanks. 

Comment 2

[Matthias Siller]

Yes, I did try all available version of Phoenix and Mozilla.

In general I can say, every version of Mozilla I've tried (and I'm using almost
every nightly build) works fine on this specific page http://derstandard.at/ for
a long time (they did a redesign of the page to make it w3c compatible and
usable for mozilla users).  

On the other hand, every version of Phoenix I've tried does not work.

This makes me think it could only be a problem in Phoenix's JavaScript handling.
Maybe Phoenix is only less forgiving on handling badly programmed JavaScript pages.

Actually I'm using:
* Mozilla: Mozilla 1.2b: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US;
rv:1.2b) Gecko/20021006
* Pheonix: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.2b)
Gecko/20021010 Phoenix/0.3

As I'm also using Linux, I'm will examine, if the bug does exist on this
platform, too.

Comment 3

[Asa Dotzler [:asa]]

Phoenix uses Mozilla's JS engine and DOM as far as I know without change. It
seems odd to me that this wouldn't work. The only differences I know of are in
pop-up blocking. 

Comment 4

[Matthias Siller]

Ok, I searched a little deeper using the JavaScript console of both programs
Mozilla and Phoenix and found a difference, as Phoenix shows errors on the
JavaScript console when opening those pages while Mozilla's JavaScript consoles
stays empty. I pasted the messages below.

******* For http://www.otb.at it logs out:

Error: uncaught exception: [Exception... "Component returned failure code:
0x8000ffff (NS_ERROR_UNEXPECTED) [nsIDOMNavigator.cookieEnabled]"  nsresult:
"0x8000ffff (NS_ERROR_UNEXPECTED)"  location: "JS frame ::
http://www.otb.at/main.htm?start=1034416738593 :: Check :: line 95"  data: no]

--> line 95 says:
if (navigator.cookieEnabled || is.opera)

******** For http://www.derstandard.at it logs out:

Error: uncaught exception: [Exception... "Component returned failure code:
0x8000ffff (NS_ERROR_UNEXPECTED) [nsIDOMNavigator.cookieEnabled]"  nsresult:
"0x8000ffff (NS_ERROR_UNEXPECTED)"  location: "JS frame ::
http://derstandard.at/js/site/State_v4.2.js :: anonymous :: line 128"  data: no]

--> line 128 says:
if (navigator.cookieEnabled && document.cookie) {

---------------------------------------------
Maybe Phoenix does not pass the function "navigator.cookieEnabled" correctly?

Comment 5

[Matthias Siller]

Ok, finally I found the cause of the problem for both pages:
Phoenix does not set the line "user_pref("network.cookie.cookieBehavior", 0);"
in prefs.js automatically. If I add this line manually Phoenix works fine.
Without this line the function "navigator.cookieEnabled" throws an error on
Mozilla an Phoenix.

Comment 6

[Shawn Lindsay]

Dropping "user_pref("network.cookie.cookieBehavior", 0);" in user.js gets rid of
the error messages.

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2b) Gecko/20021020 Phoenix/0.3,
build 2002102008

Comment 7

[David P James]

Confirming in Linux, build 20021024; changing OS to all.

I can also confirm that the user fix posted in comment #6 is a workaround.

I'm going to change the summary from:
"Phoenix JavaScript  page-replace does not work properly"

to
"Pref "network.cookie.cookieBehavior" not set automatically in Phoenix"

(see comment #4 and comment #5)

Comment 8

[Asa Dotzler [:asa]]

probably needs to be added to Phoenix's all.js

pref("network.cookie.cookieBehavior",       0); // 0-Accept,
1-dontAcceptForeign, 2-dontUse, 3-p3p

Comment 9

[Blake Ross]

*** This bug has been marked as a duplicate of 170705 ***

Comment 10

[Simon Paquet [:sipaq]]

Mass verifiying of old duplicates.