724838 - Improve layout/ui for correlations pop-up

Status: RESOLVED WORKSFORME

(Socorro :: Webapp, task)

Description

[Schalk Neethling [:espressive]]

Attachment: Correlation pop-up problems

The current view of the correlation data on top crashers has multiple problems such as content overflowing it's container, data overlapping etc.

This pop-up view should have a general make over to fix above problems and make the view generally more useful. See screenshot

Comment 1

[Robert Kaiser]

We should not waste time on this right now, we'll need to redo new UI anyhow when moving correlation reports into Socorro proper and stop using this crude cut-and-paste-from-txt approach.

Comment 2

[Schalk Neethling [:espressive]]

I cannot seem to find anywhere where the correlations actually load (on top crashers) to trigger a pop-up. Rob, is this something that is still in the works or has this functionality been removed?

Comment 3

[Robert Helmer [:rhelmer]]

(In reply to Schalk Neethling [:espressive] from comment #2)
> I cannot seem to find anywhere where the correlations actually load (on top
> crashers) to trigger a pop-up. Rob, is this something that is still in the
> works or has this functionality been removed?

It hasn't been purposefully removed, but I suspect it's broken, since I see for example https://crash-stats.mozilla.com/report/list?product=Firefox&range_value=7&range_unit=days&date=2014-09-22&signature=js%3A%3AGCMarker%3A%3AprocessMarkStackTop%28js%3A%3ASliceBudget%26%29&version=Firefox%3A32.0.2#tab-correlations which has correlations and is a topcrasher on https://crash-stats.mozilla.com/topcrasher/products/Firefox/versions/32.0.2?days=7 but there's no link on the topcrasher in the correlations column.

Comment 4

[Schalk Neethling [:espressive]]

(In reply to Robert Helmer [:rhelmer] from comment #3)
> (In reply to Schalk Neethling [:espressive] from comment #2)
> > I cannot seem to find anywhere where the correlations actually load (on top
> > crashers) to trigger a pop-up. Rob, is this something that is still in the
> > works or has this functionality been removed?
> 
> It hasn't been purposefully removed, but I suspect it's broken, since I see
> for example
> https://crash-stats.mozilla.com/report/
> list?product=Firefox&range_value=7&range_unit=days&date=2014-09-
> 22&signature=js%3A%3AGCMarker%3A%3AprocessMarkStackTop%28js%3A%3ASliceBudget%
> 26%29&version=Firefox%3A32.0.2#tab-correlations which has correlations and
> is a topcrasher on
> https://crash-stats.mozilla.com/topcrasher/products/Firefox/versions/32.0.
> 2?days=7 but there's no link on the topcrasher in the correlations column.

So, I do not see any errors in the console, it does return 0 'hits' but, in other places, such as the report list tab, I have seen it throw an error, resulting in the spinner just spinning.

I am surfacing that error in the UI as part of bug https://bugzilla.mozilla.org/show_bug.cgi?id=716497, so the user knows that the call failed and why. The most common reason I have seen is:

platform
Select a valid choice. Windows NT is not one of the available choices.

--

Are you going to open a bug to fix the correlations bit and then have this blocked on that? The one question though is, seeing that none of the users has noticed/complained, perhaps we can just remove it completely from top crashers?

Comment 5

[Schalk Neethling [:espressive]]

(In reply to Robert Helmer [:rhelmer] from comment #3)
> (In reply to Schalk Neethling [:espressive] from comment #2)
> > I cannot seem to find anywhere where the correlations actually load (on top
> > crashers) to trigger a pop-up. Rob, is this something that is still in the
> > works or has this functionality been removed?
> 
> It hasn't been purposefully removed, but I suspect it's broken, since I see
> for example
> https://crash-stats.mozilla.com/report/
> list?product=Firefox&range_value=7&range_unit=days&date=2014-09-
> 22&signature=js%3A%3AGCMarker%3A%3AprocessMarkStackTop%28js%3A%3ASliceBudget%
> 26%29&version=Firefox%3A32.0.2#tab-correlations which has correlations and
> is a topcrasher on
> https://crash-stats.mozilla.com/topcrasher/products/Firefox/versions/32.0.
> 2?days=7 but there's no link on the topcrasher in the correlations column.

Hey Rob, have you had some time to determine why the correlations on top crashers does not load? Thanks!

Comment 6

[Robert Helmer [:rhelmer]]

(In reply to Schalk Neethling [:espressive] from comment #5)
> (In reply to Robert Helmer [:rhelmer] from comment #3)
> > (In reply to Schalk Neethling [:espressive] from comment #2)
> > > I cannot seem to find anywhere where the correlations actually load (on top
> > > crashers) to trigger a pop-up. Rob, is this something that is still in the
> > > works or has this functionality been removed?
> > 
> > It hasn't been purposefully removed, but I suspect it's broken, since I see
> > for example
> > https://crash-stats.mozilla.com/report/
> > list?product=Firefox&range_value=7&range_unit=days&date=2014-09-
> > 22&signature=js%3A%3AGCMarker%3A%3AprocessMarkStackTop%28js%3A%3ASliceBudget%
> > 26%29&version=Firefox%3A32.0.2#tab-correlations which has correlations and
> > is a topcrasher on
> > https://crash-stats.mozilla.com/topcrasher/products/Firefox/versions/32.0.
> > 2?days=7 but there's no link on the topcrasher in the correlations column.
> 
> Hey Rob, have you had some time to determine why the correlations on top
> crashers does not load? Thanks!

I have not done so yet, will look today.

Comment 7

[Robert Helmer [:rhelmer]]

The problem seems to be related to the platforms being passed to the /correlations/ view:

Works:

http://localhost:5000/correlation/signatures?correlation_report_type=core-counts&product=Firefox&version=35.0a1&platforms=Windows%20NT

This is what's actually passed:

http://localhost:5000/correlation/signatures?correlation_report_type=core-counts&product=Firefox&version=35.0a1&platforms=Windows%2CMac+OS+X%2CLinux

Comment 8

[Schalk Neethling [:espressive]]

Rob, it sounded like you are actually working on this?

Comment 9

[Robert Helmer [:rhelmer]]

(In reply to Schalk Neethling [:espressive] from comment #8)
> Rob, it sounded like you are actually working on this?

Well I found the problem, I can take the bug and figure it out :)