385520 - "A Bug's Life Cycle" is misleading about WORKSFORME vs FIXED

Status: RESOLVED FIXED

(bugzilla.mozilla.org :: General, defect)

Description

[Jesse Ruderman]

https://bugzilla.mozilla.org/page.cgi?id=fields.html#status says that WORKSFORME means "All attempts at reproducing this bug were futile".  But bugzilla.mozilla.org also uses WORKSFORME when it is clear that a bug has gone away, but it isn't known why it went away (in contrast with FIXED).  This mismatch between documented meaning and actual use confuses people (e.g. bug 300982 comment 52).

I'm not sure whether this should be filed against mozilla.org:bmo or Bugzilla.

Comment 1

[Dave Miller [:justdave]]

WORKSFORME isn't the correct resolution for that situation in my opinion.  I've always advocated marking the bug as a duplicate of the bug that fixed the problem if it can be located, or resolving it as FIXED and just saying in the comment that we don't know when or how...

Now that we have better control over resolution values, maybe this should be a dupe of bug 231856?

Comment 2

[Frédéric Buclin]

(In reply to comment #1)
> WORKSFORME isn't the correct resolution for that situation in my opinion.

Why not? I use it all the time when I'm unable to find the patch which fixed the problem. Better is to fix the description of the resolution.

Comment 3

[Dave Miller [:justdave]]

(In reply to comment #2)
> (In reply to comment #1)
> > WORKSFORME isn't the correct resolution for that situation in my opinion.
> 
> Why not? I use it all the time when I'm unable to find the patch which fixed
> the problem. Better is to fix the description of the resolution.

Because it implies that the user's problem could never be reproduced.  In this case, we know it was broken and that it's been fixed, we're just not sure how or when it got fixed.  Marking it WORKSFORME is like telling the reporter that they didn't do a good enough job of providing steps to reproduce the bug, which is sort of offensive if we know the issue really did exist and no longer does.

Comment 4

[Max Kanat-Alexander]

Maybe something like FIXED_ELSEWHERE would be appropriate. Although I don't really want to have some proliferation of resolutions.

The only reason I use WORKSFORME instead of FIXED is that I like to keep track of how many patches people actually wrote for each release.

Comment 5

[:Gavin Sharp [email: [e-mail]]]

I don't think that "WORKSFORME" carries the implicit accusation of incompetence on the reporter's part. In fact in a lot of cases the reporter is the one that resolves the bug "WORKSFORME". I think it's important to distinguish "fixed by a specific code change" from "fixed, but we don't know by what"; I think it's less important to distinguish "nobody (including the reporter) can reproduce the problem anymore" from "no one can reproduce the problem and the reporter doesn't respond to requests for more details".

But either way, whether or not you think "WORKSFORME" is the correct resolution to use in the case mentioned in comment 0, the fact that it's used for that on b.m.o means we should update the documentation to reflect it's actual use.

Comment 6

[Reed Loden [:reed]]

So, somebody should write up some new documentation for me to use!

Comment 8

[石庭豐 (Seak, Teng-Fong)]

I stand in Dave's side that FIXED should be used for bugs which are *fixed* by a way or another.

Comment 9

[:Gavin Sharp [email: [e-mail]]]

That's nice, but this bug's summary isn't "determine how we should use WORKSFORME on b.m.o". That issue's been resolved for long time now. This bug is about documenting it's actual usage correctly.

Comment 10

[石庭豐 (Seak, Teng-Fong)]

Well, that is to say that I find that the current usage explanation is good enough.  No need to change.

Personally, "it works (for me)" is a polite way of mine to answer users (in my company or elsewhere) that I can't find the bug they seemingly had discovered, instead of telling them to "get lost" :)

Comment 11

[Jonathan Watt [:jwatt]]

In my opinion it is useful to be able to distinguish between bugs that were fixed by a known checkin, and bugs that disappeared for unknown reasons. As a developer and module peer I like to be able to know which bugs we genuinely did fix, and which bugs we may wish to revisit to figure out whether we should examine them further. Marking bugs that have disappeared as "fixed" breaks this, and it's silly, since we don't know whether they really were fixed.

I also strongly feel that DUPLICATE should be reserved for bugs that are logically duplicate, not bugs that happen to share a common fix. For that I feel (and we do) mark bugs as FIXED with a comment mentioning the bug that fixed the issue. Overloading terms too much is not helpful.

Comment 12

[石庭豐 (Seak, Teng-Fong)]

Actually, if they agreed on creating a new resolution, like WORKSNOW as suggested by some, I would vote for it.  But it seems this suggestion didn't please and was beaten to death.

BTW, if the term WORKSNOW is not distinguishable with WORKSFORME, how about VANISHED or DISAPPEARED?  These two don't share the same first four letters with existent resolution words.

Comment 13

[:Gavin Sharp [email: [e-mail]]]

If you think b.m.o needs a "WORKSNOW" state, please file a new bug, that's not in this bug's scope. I don't think it's a good idea, fwiw.

Comment 14

[Jonathan Watt [:jwatt]]

That's bug 231856.

Comment 15

[Dave Miller [:justdave]]

Rearranging this component to not have me as the default assignee, so that it doesn't appear like I'm intending to work on bugs that other people could be taking care of if they didn't think I was already doing them.  If this bug is a software issue on b.m.o and you'd like to fix it, the modified source is now available for the patching (see bug 373688).  Reassigning to the new default owner.

Comment 16

[Gervase Markham [:gerv]]

Fixed in bmo-4.0 as:

WORKSFORME:

Either:
a) all attempts at reproducing this bug were futile, and reading the code produces no clues as to why the described behavior would occur;
or
b) the bug was present once, but is now not reproducible (and so was probably fixed in another bug). 

Gerv

Comment 17

[石庭豐 (Seak, Teng-Fong)]

It's really a great shame that WORKSFORME has been perverted to suit some instead of having a new status being created.
OK, as you've decided to do so, then do it thoroughly and please be kind to change the description in this page as well:
https://bugzilla.mozilla.org/page.cgi?id=fields.html#resolution