884436 - Sorting by METHOD is really painful

Status: RESOLVED FIXED

(DevTools :: Netmonitor, defect, P3)

Description

[Girish Sharma [:Optimizer]]

It does sort by method, but all the GET/PUT/POST/DELETE calls keep shufflign amongst each other each time a new entry is added :/

This causes full tool invalidation and flickering and jumping and hard to locate one thing and hard to debug things scenarios.

Is that angry enough ? ;)

Comment 1

[Victor Porof [:vporof][:vp]]

We should always take timing into consideration when sorting. For example

> _byMethod: function({ attachment: first }, { attachment: second })
>   first.method > second.method,

becomes something like 

> _byMethod: function({ attachment: first }, { attachment: second })
>   first.method == second.method
>     ? first.startedMillis > second.startedMillis
>     : first.method > second.method,

Some goes with other predicates I guess. You could work on this Optimizer if you'd like. If not, I'll take a look at it at some point if nobody beats me to it.

Comment 2

[Victor Porof [:vporof][:vp]]

(although do note that I can't reproduce this, but you're probably right)

Comment 3

[Victor Porof [:vporof][:vp]]

Attachment: net-sort.patch

Comment 4

[Rob Campbell [:rc] (:robcee)]

Comment on attachment 801727 [details] [diff] [review]
net-sort.patch

Review of attachment 801727 [details] [diff] [review]:
-----------------------------------------------------------------

::: browser/devtools/netmonitor/netmonitor-view.js
@@ +620,5 @@
>      first.startedMillis > second.startedMillis,
>  
>    _byStatus: function({ attachment: first }, { attachment: second })
> +    first.status == second.status
> +      ? first.startedMillis > second.startedMillis

you could just reuse _byTiming for this and subsequent else expressions. If you didn't mind the extra send.

Comment 5

[Victor Porof [:vporof][:vp]]

https://hg.mozilla.org/integration/fx-team/rev/00e62d664a9c

Comment 6

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/00e62d664a9c