[Webkit-unassigned] [Bug 161660] New: Web Inspector: Stepping through code should point to the next thing that will execute and not skip function calls that can be stepped into

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Sep 6 17:22:22 PDT 2016


https://bugs.webkit.org/show_bug.cgi?id=161660

            Bug ID: 161660
           Summary: Web Inspector: Stepping through code should point to
                    the next thing that will execute and not skip function
                    calls that can be stepped into
    Classification: Unclassified
           Product: WebKit
           Version: WebKit Nightly Build
          Hardware: All
                OS: All
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Web Inspector
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: joepeck at webkit.org
                CC: bburg at apple.com, joepeck at webkit.org,
                    mattbaker at apple.com, nvasilyev at apple.com,
                    timothy at apple.com, webkit-bug-importer at group.apple.com

Summary:
Stepping through code should point to the next thing that will execute.

Test:
1. <script>
2. function foo(a, b) { return a + b; }
3. function a() { return 1; }
4. function b() { return 2; }
5. var x = 1;
6. foo(
7.   a(),
8.   b()
9. );

Steps to Reproduce:
1. Inspect test page
2. Set a breakpoint on line 5
3. Reload
  => Pause on breakpoint
4. Step through the code (step-over)
  => Can be confusing what is about to execute

Notes:
When stepping through code in different debugger environments we get different behavior:

  - WebKit: will stop at lines [5, 6] => "x = 1", "foo(...)"
  - Chrome: will stop at lines [5, 6] => "x = 1", "foo(...)"
  - Firefox: still stop at lines [5, 6, 7, 8, 6] => "x = 1", "foo(...)", "a()", "b()", "foo(...)"
  - lldb: would stop at lines [5, 7, 8, 6] => "x = 1", "a()", "b()", "foo(...)"

I think the lldb approach makes the most sense. It is the smartest. In WebKit/Chrome you may not realize that you could step into "a()" or "b()". In lldb it is very clear that those will execute and you are given the opportunity to step over or into them. Firefox is close, but it is unclear what the first stop on line 6 mean given the stop at 7 and 8 happen before returning to and really executing line 6.

Proposed stepping locations:

  - Proposed stop at lines [5, 7, 8, 6] => "x = 1", "a()", "b()", "foo(...)"

Reasoning:

  - Stop at any location where the developer should be offered the choice to step over or into
    - this means function calls in execution order. foo(a(), b()) stops first at a(), then b(), then foo()

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20160907/32ad280e/attachment.html>


More information about the webkit-unassigned mailing list