with NS4-Mac

changes from jake@acutex.net to make it possible to toggle the default value of newemailtech for new profiles, this is set by default to be turned on (the old default was off) ; r=dmose@mozilla.org.  changes from me to make newemailtech the default in all new installations, and update the verbiage in various spots to make it clear that newemailtech is now considered the one true way and the old tech will be going away.  r=endico@mozilla.org,cyeh@bluemartini.com

add a stack to save the current global SQL state so that queries can be nested.  used this mechanism to fix the 62618 regression.  r=encido@mozilla.org

patch from Jake <jsteenhagen@novagate.com> to fix bug 59515; adds a title attribute containing status and short description to various bug links.  In some browsers (most Mozilla builds, at least) this makes the status and short description available as a tooltip, so it is no longer necessary to click through to see it. r=dmose@mozilla.org

Bug 6682: moving location of graph creation dir to graphs instead of data/mining, for security reasons.

shown that belong to products the user can't see if usebuggroups is
active.  Patch by Matthias Radestock <rade@lshift.net>

Make 'doomed' reports honor the per-product defaultmilestone setting, rather than looking at the obsolete system-wide 'curmilestone' param.

so that saved copies of show_bug are useful. patch by dave@intrec.com, r=cyeh

The code was trying to use ConnectToDatabase to connect to the shadow db after quietly_check_login() had already called it for us under the covers _without_ the useshadowdb parameter.  So reports were being run against the main database, triggering UI stalls.  r=endico@mozilla.org

when invoked with -syncall, have the GET_LOCK time out after 45 minutes rather than 1 second, since we want syncall to happen anyway, even if other individual syncs are currently in progress.  r=endico@mozilla.org

Now it won't choke on this anymore.  Patch also includes a Perl 5.6
compatibility update (see bug 52921 and bug 44622 for details).

NS4Mac putting a single linefeed in empty fields.

patch submitted by sstock@iconnect-inc.com

large class of problems wouldn't be reported.  (It wouldn't show most
cases where a table had an invalid ID pointing into another table.)

patch by dave@intrec.com (Dave Miller)

background colors/image and link colors.

patch by JRobertson@medevolve.com (Jon Robertson)

patch by jmrobin@tgix.com

by jmrobins@tgix.com (Joe Robins). LDAP sections haven't been
tested yet, but the code is arranged such that it shouldn't disturb
existing user authentication system.