452917 - don't display the "Mark as Duplicate" link for resolved/verified bugs

Status: RESOLVED WONTFIX

(Bugzilla :: User Interface, defect)

Description

[Simon Bünzli]

How often do we expect the (verified) resolution of a bug to change due to duplication? IMHO you're currently quite encouraged to consider every bug for duplication which we haven't (needed) so far.

Removing the link when a second <select> is visible would also allow to reconsider bug 452815 which is about only seeing substantial information when scanning a bug's fields from top to bottom.

Comment 1

[Frédéric Buclin]

(In reply to comment #0)
> How often do we expect the (verified) resolution of a bug to change due to
> duplication?

This happens from time to time, while doing triage (a reporter incorrectly resolving his bug as FIXED while it's a dupe of another bug, or a previously bug marked as WONTFIX/WFM/INVALID till we realize it's in fact a dupe of another bug). But I agree this doesn't happen very often.

So what do you suggest if you still need to dupe it? Reopen + mark as dupe? I.e. a two-step process?

Note that with the previous UI, you could already mark resolved bugs as dupes, so this is nothing new here.

Comment 2

[Simon Bünzli]

(In reply to comment #1)
> So what do you suggest if you still need to dupe it?

Correct me if I'm wrong, but can't you just change the resolution from e.g. FIXED to DUPLICATE and then insert the bug number? That'd only be one additional click to what we've already got, wouldn't it?

Comment 3

[Samuel Sidler (old account; do not CC)]

(In reply to comment #2)
> Correct me if I'm wrong, but can't you just change the resolution from e.g.
> FIXED to DUPLICATE and then insert the bug number? That'd only be one
> additional click to what we've already got, wouldn't it?

That doesn't solve the JS-off case, unfortunately.

Comment 4

[Simon Bünzli]

(In reply to comment #3)
> That doesn't solve the JS-off case, unfortunately.

Guess I'm still missing something, but with JavaScript off, there's no "Mark as Duplicate" link for me, at all (resp. the link doesn't point anywhere and wouldn't work without JS anyway).

Comment 5

[Tony Mechelynck [:tonymec]]

Before, not the latest, but the previous Bugzilla upgrade, the only way to change a resolution was to first REOPEN the bug, then RESOLVE it again differently. Now it is possible to change the resolution without a separate REOPENing step, regardless of whether the current resolution is DUPLICATE, FIXED, INVALID or whatever, and regardless of which new resolution is desired. IMHO this was a great improvement, and a help to triagers. Removing the ability to change the resolution to DUPLICATE when a bug is already RESOLVED would be a step backwards on the same path.

DUPLICATE is a special resolution in that a DUPLICATE bug is a duplicate "of something". Setting the resolution to DUPLICATE also implies entering a bug number; thus, changing the resolution to DUPLICATE by means of the rolldown wouldn't be enough unless there were an additional input box "duplicate of", which might be disabled or hidden when the resolution is something else; but if we do this for changing the resolution to DUPLICATE when the bug is already RESOLVED, why not do it also when it isn't? The way I see it, _setting_ a resolution and _changing_ a resolution should happen the same way (except of course that in the former case there is a change to the status of the bug, which is a different field than the resolution itself).

Comment 6

[Simon Bünzli]

(In reply to comment #5)
> Removing the ability to change the resolution to DUPLICATE when a bug is
> already RESOLVED would be a step backwards on the same path.

Note: This is _not_ what this bug is asking for. Removing the link will _not_ remove the possibility to resolve a bug as DUPE in one swoop.

> thus, changing the resolution to DUPLICATE by means of the rolldown
> wouldn't be enough unless there were an additional input box "duplicate of"

Have you ever tried to duplicate a bug in Bugzilla 3.2? That's exactly what's happening. ;-)

The "Mark as Duplicate" link is really just a JavaScript driven shortcut for setting the bug to RESOLVED DUPLICATE and enable the field for entering the bug number. At least for UNCO and NEW bugs this can be quite handy on b.m.o, hardly for RESOLVED/VERIFIED bugs, though...

Comment 7

[Tony Mechelynck [:tonymec]]

What I'm driving at is the "principle of least surprise": why make the interface look different, when actually its capabilities are the same? It's true that an already RESOLVED bug is less often changed to DUPLICATE than a newly filed one, but as said in comment #1 the difference is only one of degree. As long as it is not totally inconceivable that a WORKSFORME or INCOMPLETE might later be found, possibly by someone else, to be a DUPLICATE of some known bug, why not leave the shortcut just like it is for UNCONFIRMED or NEW bugs?

Comment 8

[Tony Mechelynck [:tonymec]]

Maybe I spoke ambiguously. I meant: ...why not leave the shortcut there for WORKSFORME, INCOMPLETE, etc. just as for UNCONFIRMED, NEW, etc.?

Comment 9

[Max Kanat-Alexander]

It happens enough to leave the link there, and the link doesn't harm anybody.

Comment 10

[Simon Bünzli]

(In reply to comment #9)
> It happens enough to leave the link there

Do you actually have a query for all the resolved/verified bugs that later on were duped (at least for b.m.o)?

> and the link doesn't harm anybody.

Whenever I glance over a bug's status, it gets in my way - and it's really the only non-informational line in that column of the header, all other shortcuts (edit and vote) are at the end of the lines, where they don't interfere with scanning the column.

Please try using the following addition to userContent.css for a few days and then reconsider whether that wouldn't make a slightly more streamlined UI:

@-moz-document url-prefix(https://bugzilla.mozilla.org) {
  #dup_id_discoverable { display: inline !important; }
  #resolution_settings:not([style*="none"]) + * + * + #dup_id_discoverable {
    display: none !important;
  }
}

Comment 11

[timeless]

noscript:
[RESOLVED|v] [WONTFIX|v] [(duplicate)]

scripted:
[RESOLVED|v] [WONTFIX|v]
or
[RESOLVED|v] [WONTFIX|v] /duplicate/

clicking wontfix and selecting duplicate would yield:
[RESOLVED|v] [DUPLICATE|v] [              ]

I just finished suggesting this in bug 452815 which i would not have minded taking the three steps to mark. and in that time i'd have decided that they aren't really duplicates and left it as is.