Closed Bug 1406159 Opened 8 years ago Closed 8 years ago

Intermittent double page header on https://upc.ch

Categories

(Core :: Networking, defect, P3)

57 Branch
defect

Tracking

()

RESOLVED INCOMPLETE
Tracking Status
firefox56 --- unaffected
firefox57 --- fix-optional
firefox58 --- fix-optional

People

(Reporter: miketaylr, Assigned: mayhemer)

References

()

Details

(Keywords: regression, Whiteboard: [necko-triaged])

Attachments

(1 file)

Attached image double-header.png
Originally reported at https://github.com/webcompat/web-bugs/issues/11385 STR: 1. visit https://upc.ch expected: See a single "Welcome, make your selection" header actual: Sometimes, you get two. It seems to depend on a number of things, possibly location or locale. It reproduces for me and the reporter of the webcompat.com bug, but not for anyone else on my team. :/ regression range: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=10d42af6319d08daca29d5d163f301696b5ff1fd&tochange=b37a0bd71bbb1f3e5b4f58f1936d9cc0a38851d2 [Tracking Requested - why for this release]: I'm not sure how widespread this might be... but requesting tracking just in case.
Flags: needinfo?(honzab.moz)
100% repro for me locally (fortunately). I turned tailing off with network.http.tailing.enabled;false (nearly completely reverting bug 1358060) and I can reproduce as well. I confirmed that a build from a changeset just above one before bug 1358060 can't reproduce at all. Hence, one of the changes not under control of that preference must affect this somehow.
Assignee: nobody → honzab.moz
Flags: needinfo?(honzab.moz)
(In reply to Honza Bambas (:mayhemer) from comment #1) > 100% repro for me locally (fortunately). > > I turned tailing off with network.http.tailing.enabled;false (nearly > completely reverting bug 1358060) and I can reproduce as well. Only when I don't open a new tab (do a (hard)reload in the same tab.) > I confirmed > that a build from a changeset just above one before bug 1358060 can't > reproduce at all. > > Hence, one of the changes not under control of that preference must affect > this somehow.
We could use a contact at UPC to debug this. I contacted them and we'll see if we can get one of their web-devs to work with us. (If you're a UPC web-dev and you're reading this, please create a Bugzilla account and post a comment in this bug, and we'll work with you).
So, I was again once able to reproduce this with trackers tailing disabled. I start to think this is a site bug that we just expose, but I'm not 100% sure yet. The page (when the problem happens) is doing a request for [1] twice, the source script call is [2]. When the bug doesn't reproduce, this request is made only once. I don't see any immediate influence that tailing would have. When I add a breakpoint at [2], the callstack is quite long and generic. Looks like the header is added more than once to the element with ".nbx-content" class. The same element has also ".promotion-snippet" class (internally referred as LGI.NBX.SNIPPET_SELECTOR), for whose children a load is made later (being performed at [2]) that causes to show the header twice. The js code on the site is pretty complex and I'm not sure I will be able to figure this out that easily. [1] https://www.upc.ch/content/www-upc-cablecom-ch/en/teasers/default/jcr:content/promofullwidth1.html?wcmmode=disabled [2] https://www.upc.ch/etc/designs/upc-cablecom-ch/common/js/head.min.89cd65d771ff7da03c3bc4cde91f4f37.js:5303:2
Priority: -- → P3
Whiteboard: [necko-triaged]
The web page has changed (specifically now before Christmas) that it's hard to verify if the bug is still valid. Closing for now as incomplete.
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: