1047500 - Unhit debugger breakpoints reappear after reload

Status: RESOLVED INVALID

(DevTools :: Debugger, defect)

Description

[Jonathan Allard]

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.9; rv:33.0) Gecko/20100101 Firefox/33.0 (Beta/Release)
Build ID: 20140730004005

Steps to reproduce:

1. Set breakpoint
2. Reload page
3. Remove breakpoint
4. Reload page


Actual results:

Breakpoint is still there


Expected results:

Breakpoint should stay removed

Comment 1

[Jonathan Allard]

I've been getting this bug in Devtools for as long I can remember using them, make that a few months.

Comment 2

[Nick Fitzgerald [:fitzgen] [⏰PST; UTC-8]]

I did the STR with this page on bugs.js line 27 and I can't reproduce. Can you attach expand your STR with a specific webpage, source, and line number?

Comment 3

[Jonathan Allard]

I think it has to do with when the breakpoint doesn't get hit.

http://codepen.io/joallard/debug/wqKsE

STR:
1. Set breakpoint on wqKsE:21
2. Reload
3. Remove breakpoint on 21
4. Reload

=> Breakpoint is back

Comment 4

[Jonathan Allard]

Strangely, after updating to the Aug 1st Aurora, I'm getting another bug (?) that makes the breakpoint get hit even within the `if(false)` block. 

It seems that on a run which the breakpoint is hit, it will be able to be removed properly. However if the breakpoint does not get hit, it comes back on next reload.

Comment 5

[Panos Astithas (he/him) [:past] (please ni?)]

I can't reproduce the results in comment 3 in nightly. Can you try with a nightly build too, to make sure we are seeing the same thing? Also, do you get any errors in the Browser Console (Cmd-Shift-J) when you see the weird behavior?

Comment 6

[Jonathan Allard]

In comment 4, I was referring to #1051632, which makes reproducing this bug hard with the provided STRs.

I was playing with this here (http://bl.ocks.org/joallard/raw/babcee6d4024ba0cb8d8/8f8f0fe65d4983b8104c2ed6c07aa1da2227d140/), setting breakpoints in either branches. What is weird is that when I ran the file I wrote on my disk, I got all kinds of weirdness (sticky breakpoint in true branch), but I can't reproduce anything online on Blocks. And even when I download the file I can't reproduce either. I'm really scratching my head now.

I'm at a loss for a reliable STR here.


Sent from my Nightly

Comment 7

[Jonathan Allard]

However, I'm better able to reproduce the issue in Aurora 33 (Blocks URL), which means something along those lines has been fixed or at least fiddled with. Closing.