In my implementation of  issue 629964 , I came across a bug in the implementation of Browser::SetWebContentsBlocked(): it checks to see if a WebContents is the active tab of a browser window, but fails to check to see if the browser window is front in the browserlist. I started work on fixing it, but wanted to make sure I didn't regress the original issue for which that code was added.

That code turns out to be unreviewed.

It was added in r119861, and the diff is at https://chromium.googlesource.com/chromium/src/+/41f022e8bbb57b11bf5a7c71b32a656c4f1d7570%5E%21/#F0 . The function used to be called Browser::SetTabContentBlocked, and you can see that the function gained two lines at its end:

+  if (!blocked && GetSelectedTabContentsWrapper() == wrapper)
+    wrapper->web_contents()->Focus();

The review happened at https://chromiumcodereview.appspot.com/9297043 , and those two lines were not present in that review.

This was nearly five years ago, before there was a CQ, so I consider this a mistake, where a few lines of code meant to be a local patch accidentally got included in the subversion commit.

The problem here is what to do with these lines. Certainly at the time that they were written, they addressed _some_ issue that was going on, but it's unclear what that issue was, or if it's still an issue that needs addressing in this way (as opposed to a different way) or even if the applicable code is still here (this was the era of a GTK interface, and perhaps it was meant for that one).

I'm inclined to remove this code entirely and see what breaks (if anything) but I want to get thoughts.

+ some involved engineers