Opening a bug report report with a non-UCS-2 unicode character in one of the text fields (Steps to Reproduce in my case) causes all remaining input (including the offending character) to be discarded in the resulting bug report page. I'm not sure if this is appropriate here as it might be a general bugzilla bug. Reproducible: Always Steps to Reproduce: 1. Open a new bug with the FIRE character (U+1F525, see http://www.fileformat.info/info/unicode/char/1f525/index.htm under "Java Data") in the "Steps to Reproduce" text input field. Actual Results: The text input is truncated right before the FIRE character, discarding it and all remaining text. Expected Results: The bug report contains all text as entered including unicode symbols above U+FFFF.
This looks like a bug of bugzilla. See the test done on bugzilla's official testing installation : https://landfill.bugzilla.org/bugzilla-tip/show_bug.cgi?id=20289
Yes, it is a upstream bug. See https://bugzilla.mozilla.org/show_bug.cgi?id=738609