232164 - "Bugs reported Today" link on Bugzilla panel no longer works

Status: RESOLVED FIXED

(Bugzilla :: Query/Bug List, defect)

Description

[Boris Zbarsky [:bzbarsky]]

It seems that the backwards-compat stuff is not handling the "bugs filed since"
link that the bugzilla sidebar at
http://people.netscape.com/granrose/panels/bugzilla/bugzillapanel.html uses...
I'm not sure whether this was purposeful or not, but I figured I should file it.

Comment 1

[Dave Miller [:justdave]]

*** This bug has been marked as a duplicate of 232163 ***

Comment 2

[Bradley Baetz (:bbaetz)]

No, wrong bug

Comment 3

[Bradley Baetz (:bbaetz)]

This should work now, BTW. If it doesn't, add &debug=1 to the params, and paste
that output in.

*** This bug has been marked as a duplicate of 231949 ***

Comment 4

[Boris Zbarsky [:bzbarsky]]

It doesn't work.  The output is:

SELECT bugs.bug_id, bugs.bug_severity, bugs.priority, bugs.bug_status,
bugs.resolution, bugs.priority, map_assigned_to.login_name, bugs.resolution,
bugs.target_milestone, bugs.short_desc FROM bugs, profiles AS map_assigned_to
LEFT JOIN bug_group_map ON bug_group_map.bug_id = bugs.bug_id AND
bug_group_map.group_id NOT IN (9,10) LEFT JOIN cc ON cc.bug_id = bugs.bug_id AND
cc.who = 20209 WHERE bugs.assigned_to = map_assigned_to.userid AND
(bugs.bug_status IN ('UNCONFIRMED','NEW','ASSIGNED','REOPENED','RESOLVED')) AND
(bugs.product_id IN (1,6)) AND ((to_days(now()) - to_days(bugs.delta_ts)) < '2')
AND ((bug_group_map.group_id IS NULL) OR (bugs.reporter_accessible = 1 AND
bugs.reporter = 20209) OR (bugs.cclist_accessible = 1 AND cc.who IS NOT NULL) OR
(bugs.assigned_to = 20209) OR (bugs.qa_contact = 20209) ) GROUP BY bugs.bug_id
ORDER BY bugs.bug_id

Comment 5

[Dave Miller [:justdave]]

I stand by the original dupe.  The website he's pointing at is not under
mozilla.org control.  The only way WE can fix it is with the backward
compatibility code.  The bug you duped this to was specific to the link on the
index page which was changed to use the new syntax.

Comment 6

[Dave Miller [:justdave]]

*** This bug has been marked as a duplicate of 232163 ***

Comment 7

[Bradley Baetz (:bbaetz)]

The bug you duped this against was so that query.cgi displayed the query
properly after 'edit' was used.

We already have compat code for changedin - thats what bug 231949 was about.

Except that bug 231949 got fixed by fixing the query, not the compat code. so
someone needs to fix the compat code.

Comment 8

[Myk Melez [:myk] [@mykmelez]]

Attachment: patch v1: hacks a fix for problem into Search.pm

Here's some code that special-cases searches where changedin is defined,
chfieldfrom and chfieldto are not, and chfield is [Bug creation].  That's the
query on the Bugzilla panel and in Mozilla's "QA" menu.  I also simplify the
code a bit in the process of moving it around.

Comment 9

[Myk Melez [:myk] [@mykmelez]]

Note that the original code says:

push(@specialchart, ["changedin", "lessthan", $changedin + 1]);

But I do:

push(@specialchart, ["changedin", "lessthan", $changedin]);

That's because if changedin=1 you want today's bugs (AFAICT), but $changedin+1
gives you yesterday's and today's.

Comment 10

[Bradley Baetz (:bbaetz)]

Comment on attachment 139893 [details] [diff] [review]
patch v1: hacks a fix for problem into Search.pm

Hmm. I'll buy the change, but note that since 'days' is actually 'time since
midnight', you're under-reporting rather than over reporting.

I'd prefer to keep it the way it was, but r=bbaetz either way

Comment 11

[Myk Melez [:myk] [@mykmelez]]

Checking in Bugzilla/Search.pm;
/cvsroot/mozilla/webtools/bugzilla/Bugzilla/Search.pm,v  <--  Search.pm
new revision: 1.51; previous revision: 1.50
done

Note: added back the +1 after bbaetz determined that's the way it was in 2.16. 
That doesn't affect the special case, which still shows only today's created bugs.

Comment 12

[Antonis Christofides]

*** Bug 231315 has been marked as a duplicate of this bug. ***

Comment 13

[Antonis Christofides]

I had a much simpler patch ready for bug 231315, which unfortunately did not
make it for yesterday's upgrade. But it's worth considering because (except for
its simplicity) it reproduces exactly the (erroneous) behavior of the previous
version of the code, and is thus likely to break fewer existing erroneous insane
queries.

I also have a patch for the release notes there.

Comment 14

[Jon Granrose]

for the record, I'm no longer at AOL so I can't fix or remove the code on my
people.netscape.com page...