From our perspective we see only either HTML websites or we are using plugins like flash. We don't know what's behind it (like PHP, asp). There is nothing we can do without more information. Is there a public testurl ? >The user selects a file from the dialog and clicks OK. The path to the file is >returned to the Flash application. That is a file upload prompt from a flash application, right ? Firefox doesn't send any path information with HTML file upload controls. Do you tested another browser (except IE) ?
(In reply to comment #1) > From our perspective we see only either HTML websites or we are using plugins > like flash. We don't know what's behind it (like PHP, asp). Aspera Connect is behind it. http://www.asperasoft.com/en/products/client_software_2/aspera_connect_8 > There is nothing we can do without more information. > Is there a public testurl ? I'm working on it. I'll post it as soon as possible. > That is a file upload prompt from a flash application, right ? > Firefox doesn't send any path information with HTML file upload controls. > Do you tested another browser (except IE) ? No, not from Flash, from Aspera Connect. It's an active-x control. Yes, our product has been in service for three years. We test it in Windows and Mac on Safari, IE, and Firefox. This is our first new release since Firefox 3.6. It works in Internet Explorer and Safari.
>Aspera Connect is behind it. What's behind the html or plugin doesn't matter much for us. >No, not from Flash, from Aspera Connect. It's an active-x control. Firefox doesn't support activex Controls as plugin. It's either a html file upload control or something from a plugin like flash or java. Normal html form controls are working ? You can use the "add an attachment" link in this bug and click on browse. You should probably attach a stacktrace via the "add an attachment link" using this instructions: https://developer.mozilla.org/en/How_to_get_a_stacktrace_with_WinDbg (but mdc seems to be down for me at the moment)
Created attachment 457385 [details] Stack Trace from Firefox 3.6.6 I was unable to get WinDbg to run Firefox with examine child processes turned on. I followed the steps at https://developer.mozilla.org/en/How_to_get_a_stacktrace_with_WinDbg precisely. If I left that one step out, Firefox launched and I was able to continue without a problem. The output file attached displays a note "Your debugger is not using the correct symbols" that appears after I ran the !anaylze command. I downloaded the symbols from http://symbols.mozilla.org/firefox using the command .sympath SRV*c:\symbols*http://symbols.mozilla.org/firefox. The directory was created with 42 subdirectories with .pdb files in each.
(In reply to comment #3) > Firefox doesn't support activex Controls as plugin. > It's either a html file upload control or something from a plugin like flash or java. Right, my bad. There's a Aspera Web Firefox Plugin. > Normal html form controls are working ? > You can use the "add an attachment" link in this bug and click on browse. Yes. > You should probably attach a stacktrace via the "add an attachment link" using > this instructions: > https://developer.mozilla.org/en/How_to_get_a_stacktrace_with_WinDbg (but mdc > seems to be down for me at the moment) Done.
Added a URL to a test application on our extranet. I've successfully reproduced the bug in this test application. It might take 10 or 15 tries, but the browser will eventually hang.
I got different results. A windows filepicker opens, it's active (blue titlebar) but somehow doesn't seem to have focus. You can close the filepicker with the taskmanager and you can still work with Firefox after that. This is the used npapi plugin: Aspera Web for Firefox File: npasperaweb.dll Version: 188.8.131.52072 Aspera Web for Firefox
jim, more of the same
FYI: I'm only a bug triager here in bugzilla. I request additional information from the reporter that the developers need, filter out all the bugs that are not really a bug in our code and i try to reproduce the reported bugs. I'm sorry but I'm not able to produce a fix.
Adding to my summary: - the same problem happens on firefox 3.6.8 (confirmed by testing at Aspera)
@Mathias: Is there any more information we (Aspera) could provide to help your dev team work on a fix for this ?
My job here is done and only the developers (in this case email@example.com) will ask additional questions if necessary.
fixed by bug 648935, which will merge to aurora tomorrow. http://mozilla.github.com/process-releases/draft/development_specifics/