• Rebase against central again, incorporate new changes to tests

• Register onInitiatorMouseDown on the lastframe element with the link styling only

• Undo unnecessary EOF newline deletion

• Fix test failures since the cause in the initiator column has been now moved into a separate div
• Rebase against central, past Bug 1623110

• Fix test failures since the cause in the initiator column has been now moved into a separate div

• Bug 1618417 - Merge Initiator column into Cause r=Harald,Honza
• Remove link styling for the cause label in the initiator column

I felt that it would be more logical/organized if (assuming ascending order) the sorting logic sorts all rows without a stack trace before all the other rows in alphabetic order of the cause alone,
and then proceeds to sort the rest of the rows which have a stack trace and (cause) according to the filename 1st, line number 2nd ordering as I did in Bug 1592535,
because (to me) it does not make sense to be comparing a filename to a cause alphabetically when a row with only a cause represents "lack of initiator/filename".

Try run: https://treeherder.mozilla.org/#/jobs?repo=try&revision=bf15c206ae56bd4e3893c8ec00f6150c2f42a4c4

• Refactor getting the index of new and inline style sheets into different functions for readability

Try run: https://treeherder.mozilla.org/#/jobs?repo=try&revision=bf406303daed69de941d3d319720ec1af0d0895a&selectedJob=289694261

• Bug 1247083 - Numbering of inline stylesheets should count only inline stylesheets; client-side only implementation r=pbro
• Refactor the friendly index logic into a separate function in StyleEditorUI.jsm and add a createStyleSheet helper for tests

@Honza Fixed, I ran the command that was suggested in the TEST-UNEXPECTED-FAIL output of browser_webconsole_stubs_network_event.js to regenerate itself because the network event packet now has a lastFrame property.

• Regenerate the networkEvent.js stub because of the lastFrame addition to the form.

• Bug 1592535 - Add a column displaying the last frame in the stack trace info if it exists r=Harald,Honza
• Fix lint
• Make Initiator column visible by default in Nightly, add link styling to show that it is clickable
• Show ellipses when the initiator text overflows, refactor lastFrames to use ES6 Map
• Last frame in the stacktrace is at the top, not bottom
• Add initiator column mochitest based on browser_net_cause.js testing Stack Panel open on click, last frame correctly displayed, and sorted order upon column header click
• Rename "cause" to "initiator" in comment
• Ensure the widths of all the columns, including the initiator, sum up to 100%, and fix test regressions

• Rename "cause" to "initiator" in comment

• Add initiator column mochitest based on browser_net_cause.js testing Stack Panel open on click, last frame correctly displayed, and sorted order upon column header click

• Last frame in the stacktrace is at the top, not bottom

• Show ellipses when the initiator text overflows, refactor lastFrames to use ES6 Map

Many thanks for all of the feedback above. I've based the link/clickable styling on the Stack Trace pane for consistency; here is what it looks like in dark mode:

• Bug 1592535 - Add a column displaying the last frame in the stack trace info if it exists r=Harald,Honza
• Fix lint
• Make Initiator column visible by default in Nightly, add link styling to show that it is clickable

• Bug 1592535 - Add a column displaying the last frame in the stack trace info if it exists r=Harald,Honza
• Fix lint

I have written a test that checks whether the name passed into the constructor gets set on the worker after opening the debugger toolbox, and have also factored in @bhackett 's comment from here about evaluateJS potentially not returning a string. https://bugzilla.mozilla.org/show_bug.cgi?id=1589908#c13

• Bug 1589908 - The Worker name should be reflected in the threads pane r=jlast
• Return the worker name in the name getter of the WorkerTargetFront if one exists instead of in the UI, add corresponding test to browser_dbg_listworkers.js

• Omit adjacent tab checks for clarity

In D52444#1796015, @ochameau wrote:

Could you please contribute to this test in order to cover this addition?

@jlast I've rebased on top of central as of the time of writing and it plays perfectly well with the haveTabSourcesChanged and other optimizations since then

I'm not 100 percent convinced that we need 207 or 217. I get that you only want the lines in the if statement to run only if the tab moves far enough to left or right but I'm not sure it's entirely necessary. Any thoughts?

@janelledement Thanks for the comments - If these duplicated lines

moveTabBySourceId(this.draggedSource.id, targetTab);
this.draggedSourceIndex = targetTab

are not inside the if blocks, they will always fire if the tab is not moved far enough

• Bug 1568404 - Drag/drop to reorder file tabs in editor panel r=jlast
• Rebase and add comments describing the tab moving conditionals

• Add future SignalR parser file paths to the Apache 2.0 license notice instead of adding another copy of said license

Add future SignalR parser file paths to the Apache 2.0 license notice instead of adding another copy of said license Please ignore, accidentally ran arc diff against the original revision.

• Bug 1594988 - UI sub nodes in Watch expressions should not be editable r=jlast
• Update the jest snapshot since onDoubleClick has been moved from the <li/> expression container into the ObjectInspector

The one-line descriptions in JSONHubProtocol.js, IHubProtocol.js, and HandshakeProtocol.js are taken from their C# counterparts (e.g. https://docs.microsoft.com/en-us/dotnet/api/microsoft.aspnetcore.signalr.protocol.jsonhubprotocol?view=aspnetcore-3.0). There was no such description for TextMessageFormat.js since it was originally a private-only class. I couldn't find any date or email info for the license in master of their repository.

• Add licensing headers, map "type" values to their names, undo unnecessary package-lock.json change

I wonder if there is an alternative to doing the evaluation. The server should be running in the context of the worker.

Hmm, I'm afraid I couldn't figure out how / if it's possible without too much difficulty -

https://codepen.io/transfusion/pen/BaaZmWz
Expected behavior:

• Use the source ID instead of the source URL to uniquely identify the tab being moved

• Bug 1589068 - Provide MVP Signalr support for the websocket inspector r=Honza
• Bug 1589068 - Add support for HandshakeResponse, Invocation, StreamItem, and Completion incoming messages, including multiple messages in one WS frame, by including SignalR client sources r=Honza

• Fix nits and avoid wordy undefined and null checks by moving them into getters

In D50755#1540736, @davidwalsh wrote:

Is there anything we can do about this jitter? A throttle or a debounce maybe?

• Bug 1568404 - Drag/drop to reorder file tabs in editor panel r=jlast
• Improve tab reordering algorithm to avoid a feedback loop when dragging over the neighboring tab causes the current tab to lose focus and the original tab to be reselected

• Remove unused conn argument and clean up the actor in destroy()

• Bug 1528036 - Full page screenshot tool is not functional after page refresh r=yulia
• Ensure that this._onNavigate is bound in initialize in order for the destroy function to destroy the same function which was created.

Fix alphabetical order nit, extract function simulating console input, make the 2nd expression depend on the 1st, check if the input is indeed executed in editor mode, and handle the Cmd/Meta key on macOS

Summary of the process I went through to fix this bug: