This is due to http://mxr.mozilla.org/comm-central/source/mozilla/toolkit/components/places/tests/unifiedcomplete/test_remotetabmatches.js ... as thunderbird doesn't have services-sync.

Right. This test got implemented in bug 1210616 and landed 2015-11-12. Why do we run this test for TB. This is a Firefox test, is it not?

(In reply to Jorg K (GMT+1) from comment #1) > Right. This test got implemented in bug 1210616 and landed 2015-11-12. > Why do we run this test for TB. This is a Firefox test, is it not? Yes, the right thing to do is to disable this test for Thunderbird (with skip-if = appname == "thunderbird").

(In reply to aleth [:aleth] from comment #2) > Yes, the right thing to do is to disable this test for Thunderbird (with > skip-if = appname == "thunderbird"). (if there isn't something smarter along the lines of run-if = sync-enabled, that is)

Sorry for not catching that. Maybe Mark knows a better way to detect if the product supports Sync and we could just skip in the test itself.

(In reply to Marco Bonardo [::mak] from comment #4) > Sorry for not catching that. > > Maybe Mark knows a better way to detect if the product supports Sync and we > could just skip in the test itself. I think it makes sense to have a mozinfo variable for "sync" just as there already is for healthreporter, especially as there will probably be more sync work coming up that will hit the same problem. This simple patch adds a "sync" variable and a "skip-if = !sync" to that tests manifest entry.

A bit off-topic, but is there a reason the variables aren't named the same as their originals (or lower-cased equivalents)? As in this case, why "sync" instead of "MOZ_SERVICES_SYNC".

Probably just accidental, I was originally matching the existing mozinfo variables (like `os`), but then it went farther than I expected.

https://hg.mozilla.org/integration/mozilla-inbound/rev/f0c815f451f761c90634789b055090b8ea438187 Bug 1224689 - Add a 'sync' mozinfo variable and only run test_remotetabmatches.js if it is defined. r=ted