User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.2) Gecko/20040803 Build Identifier: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.2) Gecko/20040803 Mozilla 1.7.2 and Mozilla 1.6 are afflicted by a bug which, for no discernable reason, frequently results in a whole list of GUI functions failing simultaneously. The failures are: 1. Cursor stops flashing in location text box (but is still visible). 2. For any dialogs that come up with OK/Cancel buttons, the buttons change colour on clicking but nothing happens, and the dialogs cannot be closed. 3. Bookmarks drop-down menu ceases to automatically expand sub-menus when mouse hovers over them. 4. Auto-completion stops working in location text box. No drop-down box appears. 5. Typing a URL and pressing Enter sometimes has no effect i.e. nothing happens. All of these effects occur together. There have been no cases when only one or a few have occurred independently. This almost always results in the Mozilla instance becoming unusable and having to be restarted. In fact, it is just about the only reason I have to quit Mozilla. On average I experience this once a day, for a typical usage of 1 hour per day, but sometimes within just a few minutes of starting Mozilla. I have found no workaround or action that causes these behaviour problems to cease and return to normal. Because a multitude of components are affected, perhaps this bug can be diagnosed by determining the commonality between the symptoms? I realise that my "steps to reproduce" below is lacking hard detail :-) I can make stack traces and core dumps available if people want them. Reproducible: Sometimes Steps to Reproduce: I can't "reproduce" it as such but it happens so frequently that it is *the* reason why I have to restart Mozilla. 1. Open Mozilla. 2. Wait for a strange effect to occur, e.g. notice that location box cursor has stopped flashing. 3. Observe that all other noted effects are now present. Actual Results: Browser, Mail client and other components all exhibit symptoms described above. Expected Results: Not behave in this way. Only seen on MacOSX versions of Mozilla. Never seen on Win32 or Linux x86 versions.
Moving to GFX: Mac. Reason for this is that the bug described occurs when I use the default "classic" look and feel which uses native Mac OS X widgets. A friend of mine who only runs the Modern theme and has never experienced the problems I describe. Consequently I think it is highly likely to be a Mac GUI-specific issue.
(Should have added that) I've switched my own Mozilla to use Modern theme and haven't seen the bug reappear yet.
Raising to major, as this bug has now bitten me when running the "Modern" theme too. I am still having to "Force Quit" Mozilla. I'm on Mac OS X 10.3.5, latest patches, with no manual customisation I should point out. I'm surprised that no-one else has responded. Is no-one else having this problem or have all the Mac users silently deserted to Safari?
Mozilla 1.7.3 is affected as well.
*** Bug 260187 has been marked as a duplicate of this bug. ***
Reporter, after your rant in bug 260187, did it even occur to you that we're too busy ? Nobody seems to look for existing bugs anymore (there's at least a 95% probability that the bug was filed before !) Your problems seems to indicate bug 232715, but you don't mention that you were sleeping.
I don't mention I (the disk) was sleeping because I wasn't. I've seen this happen within, say, 30 seconds of opening Mozilla. Now there are two types of sleep we could be talking about. I'm assuming you're talking about OS sleep i.e. when the screen goes black. However, there is also the "disk spindown while the rest of the machine continues running and the screen stays on". I'm assuming it's not the latter because that would be a Mac OS X bug and would trash just about every app on the OS :-) (You're wrong - I did look for other bugs, yes it did occur to me etc. I spent quite a while searching for combinations of words that might match. Unfortunately "disk sleep" wasn't in my combinations.)
can you reproduce this bug with a current nightly build? with the fix from bug 232715 now landed, i wonder if it fixed the issues you mention in this bug.
I've tried, and haven't seen it, on yesterday's nightly build: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8a5) Gecko/20041002 I've done a user-forced sleep from Apple menu (kIOMessageSystemWillSleep), and a sleep in which I let the machine go to sleep after a timeout (kIOMessageCanSystemSleep), which I presume recreates the two scenarios described in bug #232715. Neither of these two has resulted in any observable bugs. WELL DONE ALL!! :-)
-> WFM per comment 9.