Rebase and squash mobile and desktop patches

The patches for mobile/android are separate from the patches for desktop. Some of these patches are similar, such as adding a mozconfig and overriding prefs. Now that Android is a first-class supported platform, we can squash some of these patches so we reduce the number of patches we need carry on top of Firefox.

Trac:
Parent Ticket: #33656 (moved)

added ReleaseTrainMigration TorBrowserTeam202006 component::applications/tor browser owner::acat parent::33656 points::1 priority::medium reviewer::gk reviewer::sysrqb severity::normal sponsor::58-must status::needs-revision tbb-9.5 type::task labels

Making a note here: ed7faf1c2b88cd025596151f2ad97892891965a4 contains:

diff --git a/toolkit/modules/AppConstants.jsm b/toolkit/modules/AppConstants.jsm
index 109f1f94e3d3..bc679fb6f9f0 100644
--- a/toolkit/modules/AppConstants.jsm
+++ b/toolkit/modules/AppConstants.jsm
@@ -387,4 +387,11 @@ this.AppConstants = Object.freeze({
 #else
     false,
 #endif
+
+  TOR_BROWSER_UPDATE:
+#ifdef TOR_BROWSER_UPDATE
+    true,
+#else
+    false,
+#endif

But TOR_BROWSER_UPDATE is used earlier in 14df73ecfdd88db89f196cecfb934dab7e552969. We should move this code block into the earlier patch.

Trac:
Keywords: N/A deleted, TorBrowserTeam201910, tbb-9.0 added
Points: N/A to 1

During the 68esr rebase, I squashed the patch for configuring about:tor as the default homepage into the branding patch.

Unfortunately, that patch changes Fennec UI behavior. We shouldn't squash Gecko modifications with Fennec UI modifications. I opened #31983 (moved) for the Fennec UI change, so we can break that functionality into a separate patch and prevent the current inconsistency on Android.

Okay, tor-browser-68.1.0esr-9.0-3 is a squashed branch that could be used for this ticket. I already started moving commits around, e.g. moving the ones needed for actually building code even before our mozconfig commit.

We should think about more changes that would facilitate bisecting and readability of the branch. E.g. moving the updater/mar signing patches earlier, grouping Mozilla patches that belong together etc.

Something we could do while working on this bug, and making the patches less desktop vs. android, is to look for pieces we disabled on desktop but not on mobile, yet we would like to have them actually disabled on both. E.g. https://bugzilla.mozilla.org/show_bug.cgi?id=1503402 comes to mind where we just no build the webcompat extension on desktop but still do so on mobile and include it there.

Another thing I am just dropping here to not forget about it: there are a number of securitySettings "fixup" commits we should squash while we are at it.

Trac:
Keywords: TorBrowserTeam201910, tbb-9.0 deleted, TorBrowserTeam201911, tbb-9.5 added

Trac:
Status: new to assigned
Owner: tbb-team to sysrqb
Cc: N/A to tbb-team

Moving tickets to December

Trac:
Keywords: TorBrowserTeam201911 deleted, TorBrowserTeam201912 added

Trac:
Keywords: TorBrowserTeam201912 deleted, TorBrowserTeam202001 added

Moving tickets to February

Trac:
Keywords: TorBrowserTeam202001 deleted, TorBrowserTeam202002 added

Trac:
Keywords: N/A deleted, ReleaseTrainMigration added

We are no longer in February, moving tickets

Trac:
Keywords: TorBrowserTeam202002 deleted, TorBrowserTeam202003 added

Trac:
Sponsor: N/A to Sponsor58

We are no longer in March

Trac:
Keywords: TorBrowserTeam202003 deleted, TorBrowserTeam202004 added

Trac:
Reviewer: N/A to acat

Trac:
Parent: N/A to #33656 (moved)

Trac:
Reviewer: acat to sysrqb
Owner: sysrqb to acat