Closed
Bug 410028
Opened 17 years ago
Closed 16 years ago
Rolling back an edit does not remove article from Waiting for review category
Categories
(support.mozilla.org :: Knowledge Base Software, task)
support.mozilla.org
Knowledge Base Software
Tracking
(Not tracked)
VERIFIED
FIXED
0.9
People
(Reporter: cilias, Assigned: ecooper)
Details
(Whiteboard: tiki_bug, tiki_upstreamed)
Attachments
(3 files)
1.96 KB,
patch
|
nkoth
:
review+
|
Details | Diff | Splinter Review |
357.36 KB,
image/png
|
Details | |
2.17 KB,
patch
|
Details | Diff | Splinter Review |
1. Go to an article edit waiting approval.
2. Go click on History.
3. Rollback to the previous version of the article.
4. Confirm.
First you'll get a blank page, but the rollback does get processed; but more importantly, the article will still appear in the out of sync category. Even if you click on Approve changes, you'll get a message saying "Approved page was last saved after most recent staging edit".
Reporter | ||
Comment 1•17 years ago
|
||
category was renamed to Waiting for review.
Summary: Rolling back an edit does not remove article from Out of sync category → Rolling back an edit does not remove article from Waiting for review category
Comment 2•17 years ago
|
||
I am unable to get a white screen. However, I notice that when a page is rolled back, the date for the new version (which is created on roll back) reflects the date of the version rolled back to rather than the time of the rollback, hence the message "Approved page was last saved after most recent staging edit". I have change this to fix this problem.
Roll backs are not supposed to remove the out of sync category. Only approves do. This makes sense because roll backs of the staging version do not affect the approved version, until the roll back itself is approved. If the approved version was also rolled back, the order of activities is to roll back approved, then roll back staging, then approve.
Status: NEW → RESOLVED
Closed: 17 years ago
Resolution: --- → FIXED
Reporter | ||
Comment 3•17 years ago
|
||
I've just tested this on <http://support-stage.mozilla.org/kb/*Thunderbird>.
The purpose of the waiting for review category is to track which articles are not the same as the live version. If an edit has been rejected, cause the staging copy to go back to being the same as the live version, there's no reason for it appear in the waiting for review category.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Comment 4•17 years ago
|
||
I see. The solution then is to add code that checks on rollback if it has been rolled back to exactly the same as the live version. If so, it will remove it from out of sync category. The check is necessary because partial rollbacks (of some but not all edits) are foreseeable.
I will get to this bug in time, as it is non-trivial. For now, you can work around by manually removing it from the category under tiki-admin_categories.php.
Updated•16 years ago
|
Assignee: nobody → smirkingsisyphus
Target Milestone: --- → 0.9
Assignee | ||
Comment 5•16 years ago
|
||
This patch seems to mimic the requested behaviour. If you rollback a staging copy to a version that matches the live copy, it will remove it from the waiting for review category.
The next time someone edits the staging copy it will be back in the review category.
Attachment #342308 -
Flags: review?(nelson)
Comment 6•16 years ago
|
||
What about the "if
you click on Approve changes, you'll get a message saying "Approved page was
last saved after most recent staging edit".
part of the bug. Is that resolved?
Reporter | ||
Comment 7•16 years ago
|
||
(In reply to comment #6)
> What about the "if
> you click on Approve changes, you'll get a message saying "Approved page was
> last saved after most recent staging edit".
>
> part of the bug. Is that resolved?
Tested on <http://support-stage.mozilla.org/en-US/kb/*Thunderbird>. If I rollback using the History page, I don't get an link to approve (i.e. re-sync with KB copy).
Comment 8•16 years ago
|
||
Comment on attachment 342308 [details] [diff] [review]
patch for removing article from review copy when rolled back to match the live version
in r19554/r19555. Please test - also test comment #6
Attachment #342308 -
Flags: review?(nelson) → review+
Reporter | ||
Comment 9•16 years ago
|
||
Yes, this is fixed. Rolling back an edit via history removes the article from the waiting for review category, and the link to approve the edit, but still keeps the rejected edit in history (bug 416422).
So instead of rollback, remove, and approve; it is rollback and remove.
Assignee | ||
Comment 10•16 years ago
|
||
Marking this as fixed as per comment 7 and comment 9.
I'm not sure how this slipped through. Stephen when you get a chance, could you verify this?
If a staging article is reverted back to the same state as its approved version, the staging article is removed from the Waiting for review category. If it has been removed successfully, you shouldn't see a link to "approve changes" after the rollback.
For reference, check comment 4.
Status: REOPENED → RESOLVED
Closed: 17 years ago → 16 years ago
Resolution: --- → FIXED
Reopening; see screenshot, which I'll attach next.
Chris and I are still seeing this, trying to roll back/remove https://support-stage.mozilla.org/en-US/kb/*Clearing+private+data.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Assignee | ||
Comment 13•16 years ago
|
||
I'm not sure what to make of that error. I tried and got an blank screen (also some sort of error).
It did however act as it should. I reverted to a copy that's the same on staging as KB, and the staging article was taken out of the waiting cat.
https://support-stage.mozilla.org/tiki-pagehistory.php?locale=en-US&page=*Clearing+private+data
I'll play with a couple others to try to figure out if my patch or something new is causing these errors.
Assignee | ||
Comment 14•16 years ago
|
||
https://support-stage.mozilla.org/tiki-pagehistory.php?locale=en-US&page=*Cannot+connect+after+upgrading+Firefox
You can see I played a little there, and everything worked as it should.
Yeah, this seems to work for me now, too; Chris, same?
Reporter | ||
Comment 16•16 years ago
|
||
<Place comment 9 here.> :-)
There's also a problem in that rolling back to a version beyond the last approved edit removes the article from the waiting for review category. I don't see a reason to rollback that far, but it may happen by accident. And the idea for the waiting for review category is to display which staging copies are not in sync with the KB counterparts. ("Out of sync" was the original name for the category.)
Assignee | ||
Comment 17•16 years ago
|
||
This patch updates the original to address comment 16. To steal a bit from IRC: "Basically, if the staging copy is the same as the KB version, it should not be in waiting for review category. If it is not the same, it should appear in the waiting for review category."
The original patch did half of this (when rollback'd to the same as a KB version), but did not modify categories if the staging copy was different after already being the same as the KB copy.
In r22113/r22114
Assignee | ||
Comment 18•16 years ago
|
||
Marking this as fixed just so we know to test this once staging comes around.
Status: REOPENED → RESOLVED
Closed: 16 years ago → 16 years ago
Resolution: --- → FIXED
Reporter | ||
Comment 19•16 years ago
|
||
Works when rolling back via the history window, but not when rolling back via the "Remove" button in the actions box.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Assignee | ||
Comment 20•16 years ago
|
||
(In reply to comment #19)
> Works when rolling back via the history window, but not when rolling back via
> the "Remove" button in the actions box.
I added bug 478685 for this. While the fix is nearly the same, it is a different problem than the original bug.
"Remove" deletes the latest revision.
Status: REOPENED → RESOLVED
Closed: 16 years ago → 16 years ago
Resolution: --- → FIXED
Verified as FIXED per the first part of comment 19, especially given that the latter half of comment 19 is taken care of by bug 478685, w00t.
Status: RESOLVED → VERIFIED
Updated•15 years ago
|
Whiteboard: tiki_bug
Comment 22•15 years ago
|
||
Code moved around a bit on both sides after this patch. While upstreaming, I found a potential issue. At some point, $staging_info is referenced, but never defined. I don't know if it's some issue during refactoring, but that part certainly does not work.
The problem is in lib/wiki/histlib.php. Easy to spot, it's the only time the variable is ever referenced.
Updated•15 years ago
|
Whiteboard: tiki_bug → tiki_bug, tiki_upstreamed
You need to log in
before you can comment on or make changes to this bug.
Description
•