lp://staging/~toykeeper/ubuntu-power-tests/apt-update

Branch merges

Propose for merging

No branches dependent on this one.

Merged into lp://staging/ubuntu-power-tests at revision 16

PS Jenkins bot: Approve (continuous-integration) on 2015-11-10

Selene ToyKeeper (community): Needs Resubmitting on 2015-11-10

Max Brustkern (community): Approve on 2015-10-21

Diff: 350 lines (+160/-32)

12 files modified

debian/changelog (+5/-0)
debian/compat (+1/-0)
debian/control (+13/-0)
debian/copyright (+22/-0)
debian/install (+1/-0)
debian/rules (+32/-0)
debian/source/format (+1/-0)
debian/tests/control (+3/-0)
debian/tests/placeholder (+2/-0)
flash/flash.sh (+14/-0)
power.py (+61/-30)
test-new-builds.sh (+5/-2)

api

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

103. By Selene ToyKeeper on 2016-05-27

Made power.py force whoopsie uploads and collect whoopsie logs,
to make sure crashes get reported and that we can later get the
correct OOPS IDs and URLs.

Also, made it bzip2 -9v the syslog+dmesg files,
since they were using so much space.

102. By Selene ToyKeeper on 2016-05-27

Added channel / revision overrides to test-new-builds.sh,
to allow for easily (re)testing old versions without fuss.
For now, it's just to help back-fill recent arale tests which
failed to produce results due to network-manager bugs.

Also, forced wifi reconnect before each test case, again to
work around network-manager bugs.

101. By Selene ToyKeeper on 2016-05-27

Spelling corrections in detect_regressions.py.

100. By Selene ToyKeeper on 2016-05-27

Removed 'or any later version' clause from license statement,
as recommended by PES.

99. By Selene ToyKeeper on 2016-05-21

Refactored detect_regressions.py to make recent new functionality cleaner.
Reformatted text report.
Added --verbose option.
Mostly, just refactoring... and lots of it.

98. By Selene ToyKeeper on 2016-05-20

Added html output to detect_regressions.py.
Made a mess of the code in the process; needs refactoring badly.

97. By Selene ToyKeeper on 2016-05-18

Oops, make it work without $DISPLAY set (like during cron).
Also, undecided about smooth vs blocky window edges, so it's
  easy to toggle now until I decide which is better.

96. By Selene ToyKeeper on 2016-05-18

Keep graph from showing anything below zero.

95. By Selene ToyKeeper on 2016-05-18

Improved power regression graphs and detection.
Switched to weighted average of previous values.
Converted expected window to block style.
Increased sensitivity of "stddev increased" detector.
Set a minimum stddev of 0.05mA to avoid triggering over nothing.

94. By Selene ToyKeeper on 2016-05-18

Added graph output option to detect_regressions.py.
Displays each build's results with stddev and expected values and
  areas flagged for statistical increases/decreases in power.