remove BT APIv1 code from settings

[Tracking Requested - why for this release]:

fixed all unit tests, will test on device and tackle rest issues

Test on Flame/512MB Master | Metric | Mean | Median | Min | Max | StdDev | 95% Bound | | --------------------- | ------ | ------ | ------ | ------ | ------ | --------- | | contentInteractive | 3235 | 3235 | 3235 | 3235 | 0 | 3235 | | navigationLoaded | 3366 | 3366 | 3366 | 3366 | 0 | 3366 | | navigationInteractive | 3366 | 3366 | 3366 | 3366 | 0 | 3366 | | startupPathEnd | 4347 | 4347 | 4347 | 4347 | 0 | 4347 | | visuallyLoaded | 4784 | 4784 | 4784 | 4784 | 0 | 4784 | | fullyLoaded | 7357 | 7357 | 7357 | 7357 | 0 | 7357 | | uss | 23.938 | 23.938 | 23.938 | 23.938 | 0 | 23.938 | | pss | 28.962 | 28.962 | 28.962 | 28.962 | 0 | 28.962 | | rss | 46.371 | 46.371 | 46.371 | 46.371 | 0 | 46.371 | With Patch | Metric | Mean | Median | Min | Max | StdDev | 95% Bound | | --------------------- | ------ | ------ | ------ | ------ | ------ | --------- | | contentInteractive | 3118 | 3118 | 3118 | 3118 | 0 | 3118 | | navigationLoaded | 3253 | 3253 | 3253 | 3253 | 0 | 3253 | | navigationInteractive | 3253 | 3253 | 3253 | 3253 | 0 | 3253 | | startupPathEnd | 4254 | 4254 | 4254 | 4254 | 0 | 4254 | | visuallyLoaded | 4909 | 4909 | 4909 | 4909 | 0 | 4909 | | fullyLoaded | 7423 | 7423 | 7423 | 7423 | 0 | 7423 | | uss | 24.352 | 24.352 | 24.352 | 24.352 | 0 | 24.352 | | rss | 46.805 | 46.805 | 46.805 | 46.805 | 0 | 46.805 | | pss | 29.395 | 29.395 | 29.395 | 29.395 | 0 | 29.395 | Though the previous performance number are decreased, The visuallyLoaded time is longer than before, needs more investigation...

The above benchmark may not accurate because I just run once, should test again with ``` raptor test coldlaunch --app clock --runs 20 ```

run 20 times on Flame/512MB, the benifit is it also saves loadtime for non-root panels. master (data from bug 1216435 comment 2) | Metric | Mean | Median | Min | Max | StdDev | 95% Bound | | --------------------- | -------- | -------- | ------ | ------ | ------ | --------- | | contentInteractive | 3666.767 | 3663.500 | 3379 | 3806 | 74.228 | 3693.329 | | navigationLoaded | 3788.533 | 3794.500 | 3509 | 3930 | 73.888 | 3814.974 | | navigationInteractive | 3788.700 | 3794.500 | 3509 | 3930 | 73.853 | 3815.128 | | startupPathEnd | 4714.900 | 4720.500 | 4497 | 4816 | 71.927 | 4740.639 | | visuallyLoaded | 5229.267 | 5220 | 5074 | 5413 | 65.699 | 5252.777 | | fullyLoaded | 7806.700 | 7802 | 7562 | 7925 | 72.935 | 7832.799 | | uss | 23.287 | 23.496 | 22.516 | 23.984 | 0.468 | 23.455 | | pss | 28.310 | 28.521 | 27.538 | 29.011 | 0.470 | 28.479 | | rss | 45.721 | 45.933 | 44.945 | 46.426 | 0.472 | 45.890 | after patch (newly tested) | Metric | Mean | Median | Min | Max | StdDev | 95% Bound | | --------------------- | -------- | -------- | ------ | ------ | ------- | --------- | | contentInteractive | 3684.500 | 3690.500 | 3377 | 3846 | 93.753 | 3725.589 | | navigationLoaded | 3808 | 3812.500 | 3535 | 3973 | 87.395 | 3846.302 | | navigationInteractive | 3808.200 | 3813 | 3535 | 3973 | 87.489 | 3846.544 | | startupPathEnd | 4763.900 | 4739.500 | 4556 | 5046 | 108.060 | 4811.259 | | visuallyLoaded | 5255.750 | 5237.500 | 5085 | 5484 | 88.934 | 5294.727 | | fullyLoaded | 7854.600 | 7839.500 | 7667 | 8048 | 90.192 | 7894.128 | | rss | 46.222 | 46.383 | 45.367 | 46.934 | 0.439 | 46.414 | | uss | 23.773 | 23.936 | 22.926 | 24.480 | 0.438 | 23.965 | | pss | 28.797 | 28.958 | 27.947 | 29.506 | 0.439 | 28.989 | The difference still not significant, so I will not do this during 2.5

Comment on attachment 8672952 [details] [gaia] gasolin:issue-1211341 > mozilla-b2g:master This patch 1. rename bluetooth_v2.html to bluetooth.html and cleanup related html tags 2. remove bluetooth v1 and version detection scripts

(In reply to Fred Lin [:gasolin] from comment #6) > run 20 times on Flame/512MB, the benifit is it also saves loadtime for > non-root panels. > > master (data from bug 1216435 comment 2) > > | Metric | Mean | Median | Min | Max | StdDev | > 95% Bound | > | --------------------- | -------- | -------- | ------ | ------ | ------ | > --------- | > | contentInteractive | 3666.767 | 3663.500 | 3379 | 3806 | 74.228 | > 3693.329 | > | navigationLoaded | 3788.533 | 3794.500 | 3509 | 3930 | 73.888 | > 3814.974 | > | navigationInteractive | 3788.700 | 3794.500 | 3509 | 3930 | 73.853 | > 3815.128 | > | startupPathEnd | 4714.900 | 4720.500 | 4497 | 4816 | 71.927 | > 4740.639 | > | visuallyLoaded | 5229.267 | 5220 | 5074 | 5413 | 65.699 | > 5252.777 | > | fullyLoaded | 7806.700 | 7802 | 7562 | 7925 | 72.935 | > 7832.799 | > | uss | 23.287 | 23.496 | 22.516 | 23.984 | 0.468 | > 23.455 | > | pss | 28.310 | 28.521 | 27.538 | 29.011 | 0.470 | > 28.479 | > | rss | 45.721 | 45.933 | 44.945 | 46.426 | 0.472 | > 45.890 | > > after patch (newly tested) > > | Metric | Mean | Median | Min | Max | StdDev | > 95% Bound | > | --------------------- | -------- | -------- | ------ | ------ | ------- | > --------- | > | contentInteractive | 3684.500 | 3690.500 | 3377 | 3846 | 93.753 | > 3725.589 | > | navigationLoaded | 3808 | 3812.500 | 3535 | 3973 | 87.395 | > 3846.302 | > | navigationInteractive | 3808.200 | 3813 | 3535 | 3973 | 87.489 | > 3846.544 | > | startupPathEnd | 4763.900 | 4739.500 | 4556 | 5046 | 108.060 | > 4811.259 | > | visuallyLoaded | 5255.750 | 5237.500 | 5085 | 5484 | 88.934 | > 5294.727 | > | fullyLoaded | 7854.600 | 7839.500 | 7667 | 8048 | 90.192 | > 7894.128 | > | rss | 46.222 | 46.383 | 45.367 | 46.934 | 0.439 | > 46.414 | > | uss | 23.773 | 23.936 | 22.926 | 24.480 | 0.438 | > 23.965 | > | pss | 28.797 | 28.958 | 27.947 | 29.506 | 0.439 | > 28.989 | > > > The difference still not significant, so I will not do this during 2.5 Does this mean that we are actually slightly regressing the cold start perf?

I mean the first 6 metrics

Comment on attachment 8672952 [details] [gaia] gasolin:issue-1211341 > mozilla-b2g:master This is great, Fred! Just a couple of nits in the PR

The test is still in same loadtime level. And in reviewed patch I've removed some promise/lazyload parts, so the performance should be better or on par with previous test.

Issue addressed, treeherder green, merged https://github.com/mozilla-b2g/gaia/commit/cec4c1d3729137a24163756d15f98b0d37803966 thanks!

Seems loadtime in good shape https://raptor.mozilla.org/dashboard/script/measures?var-device=flame-kk&var-memory=512&var-branch=master&var-test=cold-launch&panelId=11&fullscreen