there is an non-stopping ramble about the complexity of access keys and
how to detect conflicts there in the l10n community.
So I started to grep for accesskey and the best hook I found to look for
conflicts would be the #ifdef DEBUG_jag part in
http://lxr.mozilla.org/seamonkey/source/content/events/src/nsEventStateManager.cpp#4705
Is that true? Would it be totally evil to make that a
if (sDEBUGAccessKey)
and have a hidden pref to trigger that code? Then we could set that on
and report conflicts to the error console.
I'd be happily suprised if there was a simpler and less invasive way to
test for conflicting access keys. My naive "walking-the-DOM" attempt
failed instantly.
Axel