182184 - throbber/back/forward stopped working on 20021126

Status: VERIFIED FIXED

(SeaMonkey :: UI Design, defect)

Description

[Mitch]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3a) Gecko/20021127
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3a) Gecko/20021127

The throbber has stopped throbbing when browsing using the tabbed
functionality with mozilla cvs 20021127. Works fine on a single (non
tabbed) config.

Reproducible: Always

Steps to Reproduce:
1. Start browser, open any page, click on a link
2. Open new tab (Cntrl-T), open a www page, click on link
3.  
Actual Results:  
Throbber not throbbing

Expected Results:  
Throbber throbbing

Worked on yesterdays cvs build i believe 20021126

Comment 1

[Frederic Bezies]

Confirming this bug, using a 3 hours old CVS based build - WinXP-Sp1.

It is very strange to see. Moving OS to All, as I see it on WinXP-Sp1.

Comment 2

[Andrew Schultz]

also busted with linux trunk build 2002112621 (2002112605 (CVS) works fine)

JS Console says:
Error: NS_NET_STATUS_READ_FROM is not defined
Source File:
chrome://global/content/bindings/tabbrowser.xml#tabbrowser.mTabProgressListener()
Line: 89

The icon for the tab is also never updated to the "busy" or favicon state.

This appears to be a regression from bug 181689

==> XP/Gui

Comment 3

[Andrew Schultz]

reassign

Comment 4

[Tracy Walker [:tracy]]

I am seeing this on commercial trunk builds whether there is tabbed browsing or not.
as seen on builds:

windows 2002-11-27-04-trunk
linux 2002-11-27-04-trunk
mac os9 2002-11-27-03-trunk
mac osx 2002-11-27-03-trunk

Comment 5

[Andrew Schultz]

yes.  and I get the JS error from comment 2 even without tabbed browsing

Comment 6

[Andrew Schultz]

*** Bug 182246 has been marked as a duplicate of this bug. ***

Comment 7

[Andrew Schultz]

upping severity again.  so much bustication.

Comment 8

[Frank Wein [:mcsmurf]]

On Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.3a) Gecko/2002112607 it
works

Comment 9

[[e-mail]]

Attachment: Proposed patch

Comment 10

[John Morrison]

Sorry about breaking this (and not noticing in testing before checkin).
I (stupidly) searched on the first const value in lxr, not the second and
third.

Comment 11

[Christopher Aillon (sabbatical, not receiving bugmail)]

Comment on attachment 107621 [details] [diff] [review]
Proposed patch

Haha, I did the same.  My fault also for trusting your search results.

Comment 12

[John Morrison]

Um, but actually, can't we just remove the test of aStatus in tabbrowser.xml?
Isn't nsBrowserStatusFilter.cpp already refusing to forward these notifications?

http://lxr.mozilla.org/seamonkey/source/xpfe/browser/src/nsBrowserStatusFilter.cpp#117
http://lxr.mozilla.org/seamonkey/source/xpfe/browser/src/nsBrowserStatusFilter.cpp#236

(... "fool me once, shame on you; fool me twice, shame on you" ... :-)

Comment 13

[John Morrison]

Um, scratch that. These notifications do come through on occasion.

Comment 14

[John Morrison]

Um, scratch the "scratch that". I confused my debugging output and mixed up
the printf in nsBrowserStatusFilter.cpp with the dump I put in tabbbrowser.xml.

So, it does appear that we can remove the test of aStatus in tabbrowser.xml.
Is that right, jag? darin?

Comment 15

[Felix Miata]

OS/2 trunk. Broken in 2002112712. OK in 2002112612. Using modern w/ text only icons.

Sometimes back button works. Forward seems to be staying greyed.

Comment 16

[Felix Miata]

After I hit submit on last comment and switched to another tab, fwd ungreyed and
back (appropriately for that tab) went grey. Status bar is staying blank.

Comment 17

[Felix Miata]

Both directions seem to work since I closed the tab with that last frames page.

Comment 18

[Darin Fisher]

Comment on attachment 107621 [details] [diff] [review]
Proposed patch

>+          const NS_ERROR_MODULE_NETWORK = 2152398848;
>+          const NS_NET_STATUS_READ_FROM = NS_ERROR_MODULE_NETWORK + 8;
>+          const NS_NET_STATUS_WROTE_TO  = NS_ERROR_MODULE_NETWORK + 9;

i'm more of a fan of hex when speaking of nsresult values.  0x804b0000
is much more easily recognizable than 2152398848.

also, it really bites to have to hardcode these values in JS.

sr=darin w/ the switch to hex.

Comment 19

[Hixie (not reading bugmail)]

This makes the browser unusable. Upgrading to smoketest blocker which it really
should have been from the start.

Comment 20

[John Morrison]

fixed, including switch to hex. I filed bug 182398 because I think we don't 
even need to test against those consts.

Comment 21

[Kalle Valo]

*** Bug 182249 has been marked as a duplicate of this bug. ***

Comment 22

[R.K.Aa.]

*** Bug 182501 has been marked as a duplicate of this bug. ***

Comment 23

[Daniel Yanez]

It's possible to modify the summary in order to include last 2 dupes (none talk
about throbber or buttons) and avoid more dupes in the future?

Comment 24

[Max Alekseyev]

*** Bug 182244 has been marked as a duplicate of this bug. ***

Comment 25

[Tracy Walker [:tracy]]

verfied fixed as seen on:

windows 2002-12-02-04-trunk
linux 2002-12-02-04-trunk
mac os9 2002-12-02-03-trunk

Comment 26

[jag (Peter Annema)]

As far as I can tell we should just be able to remove those filters now. Let me
take another look though.

Comment 27

[Daniel Wang]

*** Bug 182266 has been marked as a duplicate of this bug. ***

Comment 28

[Daniel Wang]

*** Bug 182970 has been marked as a duplicate of this bug. ***

Comment 29

[Daniel Wang]

*** Bug 182970 has been marked as a duplicate of this bug. ***