Open Bug 621308 Opened 14 years ago Updated 2 years ago

Tab-modal prompts don't play well with aborting the document load

Categories

(Toolkit :: General, defect)

defect

Tracking

()

Tracking Status
blocking2.0 --- -

People

(Reporter: kdevel, Unassigned)

References

Details

(Keywords: regression)

User-Agent:       -
Build Identifier: Mozilla/5.0 (X11; Linux x86_64; rv:2.0b9pre) Gecko/20100101 Firefox/4.0b9pre

Follow up to Bug 619644 and Bug 619645 (possibly same underlaying cause).

Reproducible: Always

Steps to Reproduce:
1. open reduced testcase https://bugzilla.mozilla.org/attachment.cgi?id=498236 
2. press stop-button with mouse (stop loading page)
3. press ESC key -> 
   a) original content vanishes
   b) spinner spins
Actual Results:  
1. Original content vanishes.
2. Spinner spinns.

Expected Results:  
Behave as if "cancel" has been pressed right after step 1, i.e.
1. Keep Original content.
2. Stop spinning.
3. Display Extra Content
Status: UNCONFIRMED → NEW
Ever confirmed: true
blocking2.0: --- → ?
OS: Other → All
Product: Firefox → Toolkit
QA Contact: general → general
Hardware: Other → All
Summary: Tab spinner persists after pageload complete iff first stop button pressed and then javascript confirm prompt closed with ESC (and script contains document.write → Tab-modal prompts don't play well with aborting the document load
Version: unspecified → Trunk
So dolske and I discussed this a bit on IRC and we're not sure what the expected behavior is here.  Maybe bz has some thoughts?
I'm not sure what the question is.  The behavior should be identical to the script making a stop() call at whatever point you hit the stop button, no?
OK, so, if the behavior is supposed to be identical to window.stop(), that means we need to abort the currently running script, right?
Hmm, it's more complicated than that.  We ignore document.write after a window.stop call, but not window.alert calls.
Calling stop() doesn't do anything about running scripts.
Blocks: 59314
Keywords: regression
Quite an edge case and it doesn't seem like we know what we should do here. Re-nom if there is something nasty I'm missing.
blocking2.0: ? → -
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.