hello chris, is the crash reproducible by a certain action on the website? if so, i'd recommend filing a bug for it at bugzilla.mozilla.org.

Yes, it is easily reproducible, but the website requires an account to access the functionality that fails. We may be able to provide a demo account for Mozilla, but it can not be provided to the public. How should I proceed?

Thanks! Chris

in this case you could hand out these test account credentials on a need-to-know basis via email and add say so in your original bug report (you don't have to provide a mail address publicly, it will be visible to users registered at bugzilla). if unsure, file it in the firefox > untriaged section.

if you want to, you can send me credentials via PM, then i could have a look at it and try to pin down the regression to a certain date/changeset in firefox 31. this will make live easier for developers and possibly make a bug report more actionable from the beginning...

Getting permissions for a demo account and setting it up is a lengthy process for us. We have a found a behavioral work around in the code that resolves the issue.

We do believe there is an underlying bug in Firefox, and hopefully the stack trace in the crash report will be enough for your dev team to address it.

Thanks for your help!

thanks for reporting back - in case the issue can be reduced to a simple testcase which reproduces a crash & that could also be published, this would still be immensely helpful. i don't think this issue is currently on the radar for developers...