We don't really need to look for fragments that are pulled in by [% INCLUDE %] or [% PROCESS %].  While removing this code bit doesn't allow us to seperatly check that those fragments exist and compile, they'll be checked atomatically when the the template that wants them is run through the process() routine by the 004template.t test.  This issue was raised because bug 98707 introduced a [% BLOCK %] element and the syntax for using that is the same as for including a template fragment.

Fix license headers and make attachment changed page look like bug changed page. Bug 99518 and 98110. r-myk, no second review.

Fix for bug 86300 - If a bug didn't exist and GetBugLink() tried to create a tooltip for it, you'd get uninitialized variables warnings in your error log.  This path also introduces a cache so if the same bug # is mentioned more than once during the same running of the script, it only has to query the database once.
r= mattyt, gerv

r=justdave,gerv

r=bbaetz,dkl

Patch by Myk Melez <myk@mozilla.org>.
r=bbaetz,gerv

Fix for bug 99519: timestamps were not being set correctly in the activity table in some situations, and the delta_ts on the
bug itself was not always being updated if dependencies or CCs changed.
Patch by Dave Miller <justdave@syndicomm.com>
r= bbaetz, jake

Fix for bug 104652: Duplicate bugs in the dependency tree now get marked with the message "This bug appears elsewhere in this tree." so users know why the bug does not appear to have dependencies.
Patch by Gerv <gerv@mozilla.org>.
r=jake@acutex.net,myk@mozilla.org

Patch by Jake <jake@acutex.net>.
r=bbaetz,myk

Patch by Jake <jake@acutex.net> and Bradley <bbaetz@cs.mcgill.ca>.
r=jake,myk for Bradley's portion, r=bbaetz,myk for Jake's portion.

Patch by Jake <jake@acutex.net>
r=bbaetz,myk

fix to bug 107718. This should be removed when the group stuff lands (bug
60822).

r=justdave x 2

the group

r=justdave,jake

Patch by Christopher Aillon <caillon@returnzero.com>
r= timeless@bemail.org, kiko@async.com.br

Refixing bug 71840 - The name should be the 'name' should be the same as what the 'href' points it too...
r= justdave@syndicomm.com

SECURITY FIX see bug 108385: Due to trusting of passed form fields that shouldn't have been trusted, it was possible to add a comment to a bug pretending to be someone else if you edited the HTML by hand before submitting.  The bug form did not include the field in question, but due to legacy processing code, the field was still trusted if it was present.
Patch by Dave Miller <justdave@syndicomm.com>
r= jake x2

SECURITY FIX see bug 108516: Due to trusting of passed form fields that shouldn't have been trusted, it was possible to file a
bug pretending to be someone else if you edited the HTML by hand before submitting.
Patch by Dave Miller <justdave@syndicomm.com>
r= bbaetz, jake

r=bbaetz, jake
Windows XP references are #'d because of bug 92763

bug in the list.

Do bit fiddling instead of adding groupsets from the first bug.

r=justdave, jake

Patch by Adam Kennedy <bugzilla@ali.as>
r= justdave x2

Patch by Adam Kennedy <bugzilla@ali.as>
r= justdave x2

intentional!

No review needed for tests at this time.

Patch by Myk Melez <myk@mozilla.org>.
r=gerv@mozilla.org,jake@acutex.net

checksetup.pl use the emailregexp set in params if the params file exists, or ensures that it matches the
default emailregexp from defparams.pl.

r1=justdave, r2=bbaetz.

 * If there's a compilation error, report what it is
 * Don't try to compile a template if it doesn't exist
   - We already tested for that and issued an ERROR
 * Define the 'url' FILTER

Patch by Matthew Tuck <matty@chariot.net.au>
r= dkl@redhat.com, jake@acutex.net

Bug 81594 - SQL error after editing user entry when changing numerous things at once (including the login name).
Patch by Matthew Tuck <matty@chariot.net.au>
r= dkl@redhat.com, jake@acutex.net

Patch by Myk Melez <myk@mozilla.org>.
r=gerv@mozilla.org,jake@acutex.net

Bug 104065 - Stop uninitilized string warnings from getting into the error log when the login cookie doesn't exist.
Patch by Dave Miller <justdave@syndicomm.com>
r= gerv@mozilla.org, jake@acutex.net

warning was hidden from the tests by bug 106424.

r=Jake, no 2nd review needed.

Don't rely on the TEST_VERBOSE environment variable (no longer exported from runtests.sh) and instead print to the TESTOUT file handle pulled in from Test::More.  This will allow the testing backend to check for verbosity rather than having to handle it in the .t files.

Bug 106424 - We weren't going orange on warnings such as "used only once".  This script now properly outputs the --WARNING and fails on such a condition.

Patch by Dave Miller <justdave@syndicomm.com>.
r=myk@mozilla.org, no second review needed.

Also fixes seeing bugs in the buglist (bug 95024), dependancy lists,
tooltips, duplicates, and everywhere else I could see which checked group
bugs.groupset == 0.

Also fxed bug 101560, by clearing BASH_ENV

r=myk,justdave

Bug 63249 - The Bug Counts report was running very slowly due to unneeded fields/joins in the SQL query.
Patch by Matthew Tuck <matty@chariot.net.au>
r= gerv@mozilla.org, jake@acutex.net

Patch by Gerv <gerv@mozilla.org> and Myself.
r= jake@acutex.net, gerv@mozilla.org, justdave@syndicomm.com