changed-by-other-than, or last-changed-by?
Aaron Larson
week. The CCB occasionally changes a bug during the review. The next
week they'd like NOT to see bugs where there has been no activity
since the last review. However, people are free to make bug updates
during the review.
In our legacy system, they did a query something like "show bugs
changed in the last 8 days where the 'last changed by' was something
other than the CCB review user".
I'm coming up blank on how to do this same thing in Bugzilla. Any
suggestions?
I thought that adding a search entity for "last changed by" analogous
to "last changed date" might be the trick, but I was hoping someone
would have an easier (i.e., already implemented) solution.
Joel Peshkin
You use 2.20 or newer and use the boolean charts. You use a chart that
searches for change after xxxxx AND changeby notequal ccb-user.
Aaron Larson
JP> Aaron Larson wrote:
>> We have a change control board (CCB) that reviews bug activity once a
>> week. The CCB occasionally changes a bug during the review. The next
>> week they'd like NOT to see bugs where there has been no activity
>> since the last review. However, people are free to make bug updates
>> during the review.
>> In our legacy system, they did a query something like "show bugs
>> changed in the last 8 days where the 'last changed by' was something
>> other than the CCB review user".
>> I'm coming up blank on how to do this same thing in Bugzilla. Any
>> suggestions?
>> I thought that adding a search entity for "last changed by" analogous
>> to "last changed date" might be the trick, but I was hoping someone
>> would have an easier (i.e., already implemented) solution.
JP> You use 2.20 or newer and use the boolean charts. You use a chart
JP> that searches for change after xxxxx AND changeby notequal ccb-user.
That query is what I need, but I don't how to do it with the 2.20
search page:
http://landfill.bugzilla.org/bugzilla-2.20-branch/query.cgi
There is no 'changeby' field in the leftmost dropdown in the boolean
charts.
Perhaps you were thinking of something like:
(not (<field xxx> changedby ccb))
but I don't think that would work because a bug would have been
changed by CCB, what I'm looking for is "has it been changed by
someone else" (preferably with the clause "since it was last changed
by CCB", but that is not required, just desired).
Joel Peshkin
Yeah, you're right. You can do it for comments or for any one field,
but not all fields. It would make a good RFE. Perhaps we should tell
it that "Last Changed Date" "changedby" should mean "anything"
"changedby" ???
Go ahead and file a bug. If I get inspired, (like if I find my own need
for it) I'll create a patch.
Aaron Larson
AL> We have a change control board (CCB) that reviews bug activity once a
AL> week. The CCB occasionally changes a bug during the review. The next
AL> week they'd like NOT to see bugs where there has been no activity
AL> since the last review. However, people are free to make bug updates
AL> during the review.
AL> In our legacy system, they did a query something like "show bugs
AL> changed in the last 8 days where the 'last changed by' was something
AL> other than the CCB review user".
AL> I'm coming up blank on how to do this same thing in Bugzilla. Any
AL> suggestions?
AL> I thought that adding a search entity for "last changed by" analogous
AL> to "last changed date" might be the trick, but I was hoping someone
AL> would have an easier (i.e., already implemented) solution.
JP> You use 2.20 or newer and use the boolean charts. You use a chart
JP> that searches for change after xxxxx AND changeby notequal ccb-user.
AL> That query is what I need, but I don't how to do it with the 2.20
AL> search page:
AL> http://landfill.bugzilla.org/bugzilla-2.20-branch/query.cgi
AL> There is no 'changeby' field in the leftmost dropdown in the boolean
AL> charts. Perhaps you were thinking of something like:
AL> (not (<field xxx> changedby ccb))
AL> but I don't think that would work because a bug would have been
AL> changed by CCB, what I'm looking for is "has it been changed by
AL> someone else" (preferably with the clause "since it was last changed
AL> by CCB", but that is not required, just desired).
JP> Yeah, you're right. You can do it for comments or for any one field,
JP> but not all fields. It would make a good RFE. Perhaps we should tell
JP> it that "Last Changed Date" "changedby" should mean "anything"
JP> "changedby" ???
Perhaps I'm being dense here, but if my query is;
summary changed by ccb
Say that CCB changed the summary, then a day later Aaron changes the
summary. I presume that the query will succeed. The negation of the
query will (presumably?) fail. In other words "not changed by" is
different than "changed by someone other than".
My sense is there is no way to do what I want to do, sort of
generating a query which is the cross product of
"all fields" changedBy "all users other than CCB"
It would seem like there would be a similar issue with any sort of
query were the set of potential matches is a set (perhaps this is the
only case). It seemed odd to me that someone else hadn't had this
issue before, and I figured I was just lost.
JP> Go ahead and file a bug. If I get inspired, (like if I find my own
JP> need for it) I'll create a patch.
I'm not sure what the general desired result would be. It just
doesn't seem like the current approach fits without some more
prediates, and even then the negation, while well defined, is fairly
counter intuitive.
"Last changed by" as a "field" would work for me, but I'm not sure its
what is most useful. I also didn't know how to construct an SQL query
that would satisfy it (although that particular ignorance I can solve).
Joel Peshkin
> query were the set of potential matches is a set (perhaps this is the
> only case). It seemed odd to me that someone else hadn't had this
> issue before, and I figured I was just lost.
>
> JP> Go ahead and file a bug. If I get inspired, (like if I find my own
> JP> need for it) I'll create a patch.
>
> I'm not sure what the general desired result would be. It just
> doesn't seem like the current approach fits without some more
> prediates, and even then the negation, while well defined, is fairly
> counter intuitive.
>
> "Last changed by" as a "field" would work for me, but I'm not sure its
> what is most useful. I also didn't know how to construct an SQL query
> that would satisfy it (although that particular ignorance I can solve).
Essentially, nobody ever "changes" the "LastChanged" field. So, when
someone asks about an acitivity of "who changed" the "Last changed"
field, we should (and can) re-interpret it as who made a change to any
field.
Anyway, go ahead and file an Enhancment request on bugzilla.mozilla.org
and we'll eventually do something about it.