It appears however, that the original FAQ identifies this as a bug in
mozilla's jsd code - not in firebug. From the comment at
https://bugzilla.mozilla.org/show_bug.cgi?id=307984#c34, it appears
that firebug has a workaround in place for that issue - is that the
case?
BTW - in order to get a trac account for bugs.dojotoolkit.org, you
just need to create an account at www.dojotoolkit.org. The same
account is used in both places.
-Nathan
On Nov 3, 12:46 pm, toonetown <nathan.to...@gmail.com> wrote:
> It appears however, that the original FAQ identifies this as a bug in
> mozilla's jsd code - not in firebug. From the comment athttps://bugzilla.mozilla.org/show_bug.cgi?id=307984#c34, it appears
> that firebug has a workaround in place for that issue - is that the
> case?
> It appears however, that the original FAQ identifies this as a bug in
> mozilla's jsd code - not in firebug. From the comment athttps://bugzilla.mozilla.org/show_bug.cgi?id=307984#c34, it appears
> that firebug has a workaround in place for that issue - is that the
> case?
Mozilla's jsd code does not support eval() debugging. Specifically it
assigns the caller's source file name to the eval() buffer, so the
line numbers from the eval() overlap the line numbers of the caller.
In Firebug I fix this by breakpointing on every script compilation,
examining the call stack to determine that eval() is involved,
computing an MD% hash of the eval() buffer, and using the hash to
track the buffer for debugging.