First click anywhere in the JSON viewer scrolls back to the top
Categories
(DevTools :: JSON Viewer, defect)
Tracking
(firefox-esr68 unaffected, firefox-esr78 wontfix, firefox76 wontfix, firefox77 wontfix, firefox78 wontfix, firefox79 fixed)
People
(Reporter: zbraniecki, Assigned: yzen)
References
(Regression)
Details
(Keywords: regression)
Attachments
(3 files, 1 obsolete file)
STR:
- Open https://bug1641719.bmoattachments.org/attachment.cgi?id=9152582 in the JSON Viewer
- Scroll a bit
- Click on any item below the fold
Current result:
The view gets scrolled to the top.
Expected result:
The element expands.
This only happens on the first click. After that, next click expands rows as expected.
Reporter | ||
Comment 1•5 years ago
|
||
Comment 2•5 years ago
|
||
So when you load it, no row is selected initially.
When you mousedown, if there is no selected row, the 1st one is selected for some reason.
When a row is selected, it's scrolled into view.
Then, if it was already scrolled to the top, and you don't move the mouse, the mouseup will happen in the same element, which will produce a click, which will select the desired row.
Otherwise, the mouseup will most probably happen in some other element, which won't produce a click, so the 1st row stays selected.
Reporter | ||
Comment 3•5 years ago
|
||
Harald Kirschner :digitarald
ah, found it
zbraniecki: https://searchfox.org/mozilla-central/source/devtools/client/shared/components/tree/TreeView.js#315 is called for "focus"
events fired are focus-click-focus-focus, so first focus does not have a selected state yet in the Tree I guess
Reporter | ||
Comment 4•5 years ago
|
||
Another STR:
- Visit https://api.github.com/
- Click on any node
Current Result:
The first node is selected for a split second before the clicked node gets selected.
Expected Result:
Only the clicked node should get selected.
Reporter | ||
Comment 5•5 years ago
|
||
Kwan
regression window is currently https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=d5efb485c09c230a90fc2a24ba473ae35c1963f2&tochange=ef8f91e97c99d3013dbe78e24f8cb52765ca6262 so I'm assuming Harald is right about 1545727 being the regressor
Reporter | ||
Comment 6•5 years ago
|
||
Comment 7•5 years ago
|
||
18:00.25 INFO: Narrowed integration regression window from [903f141a, 41f8b0b2] (3 builds) to [7192ec5b, 41f8b0b2] (2 builds) (~1 steps left)
18:00.25 INFO: No more integration revisions, bisection finished.
18:00.25 INFO: Last good revision: 7192ec5bf245b12aead3067b33e80fa8614a5870
18:00.25 INFO: First bad revision: 41f8b0b21560997f8b63b0389afbce68235cdced
18:00.25 INFO: Pushlog:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=7192ec5bf245b12aead3067b33e80fa8614a5870&tochange=41f8b0b21560997f8b63b0389afbce68235cdced
Updated•5 years ago
|
Updated•5 years ago
|
Comment 8•5 years ago
|
||
Too late to fix in 77, wontfix for this release.
Comment 9•5 years ago
|
||
I can't seem to add any comments in the linked PR. I get this error:
Members of the project "Restricted Project" can take this action.
Anyways, I wanted to say that I did try the approach in that PR when implementing keyboard selection - to always ensure that something always has focus, but there were 2 concerns raised:
- Will it accidentally move keyboard focus?
- This could be confusing visually unless a separate set of styles is used for when something is selected vs actually having keyboard focus if the first row always gains focus on initial render.
Reporter | ||
Comment 10•5 years ago
|
||
Comment 11•5 years ago
|
||
The severity field is not set for this bug.
:Honza, could you have a look please?
For more information, please visit auto_nag documentation.
Assignee | ||
Comment 12•5 years ago
|
||
Hi Zibi,
I just tested with only https://phabricator.services.mozilla.com/D79322 applied (to master) and it seems to work as expected (e.g. no scroll). I'm attaching a video with what I see, let me know if my steps are different from yours.
Updated•5 years ago
|
Reporter | ||
Comment 13•5 years ago
|
||
Ah, you're right! It does fix it! Thank you!
I abandoned my revision and handing the bug over to you. I think your patch is sufficient!
Reporter | ||
Updated•5 years ago
|
Assignee | ||
Comment 14•5 years ago
|
||
(In reply to Zibi Braniecki [:zbraniecki][:gandalf] from comment #13)
Ah, you're right! It does fix it! Thank you!
I abandoned my revision and handing the bug over to you. I think your patch is sufficient!
Great!
Updated•5 years ago
|
Comment 16•5 years ago
|
||
Comment 17•5 years ago
|
||
bugherder |
Updated•5 years ago
|
Description
•