Filing bug on this build warning: { /mozilla/content/base/src/nsEventSource.cpp: In member function ‘nsresult nsEventSource::SetReconnectionTimeout()’: /mozilla/content/base/src/nsEventSource.cpp:1113:35: warning: comparison of unsigned expression >= 0 is always true [-Wtype-limits] } for this assertion: > 1113 NS_ASSERTION(mReconnectionTime >= 0, "mReconnectionTime lies"); where mReconnectionTime is unsigned (and has been ever since it & this assertion were introduced in bug 338583): http://hg.mozilla.org/mozilla-central/rev/a3903f835a81#l6.152 http://hg.mozilla.org/mozilla-central/rev/a3903f835a81#l5.1027 Looks like the assertion can just go away.

(this is a warning in GCC 4.6.3, FWIW - I think that's the first version of GCC to warn about this)

Note: AFAICT, this value is read from an about:config pref, and then dynamically updated later on and clamped to the range [MIN_RECONNECTION_TIME_VALUE, MAX_RECONNECTION_TIME_VALUE] So, it doesn't look like we're particularly worried about it underflowing or anything.