►
From YouTube: Development Metrics Working Group 2019-06-27
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
A
A
A
A
It
could
be
correlating
with
with
your
hypothesis
Craig,
because
the
memory
mods
are
going
up
with
it.
More
people
are
coming
in,
so
the
division
factor
is
more
hen
CI
which
is
coming
down,
but
this
is
great
I.
Think
we're
going
up
from
trajectory
and
I'm
really
happy
to
see
this
and
yeah
the
with
community
contributions.
That's
also
telling
the
same
story
at
1700
great,
an
all-time
high
with
community
contributions
it's
1700
also,
but
on
the
on
the
topic
of.
C
No,
nothing,
nothing
to
add.
It
is
definitely
one
of
the
things
I
like
to
do
is
add
more
debugging.
You
know
numbers
so
number
of
contributor
against
each
throughput
would
be
great
because
it'll
give
us
some
insight
into.
Is
that
the
factoring
divider
that's
causing
the
number
to
like
the
our
average
to
go
down
and
so
on,
but
yeah?
No
sorry
I
need
to
dig
in
a
bit
more
and
and
figure
out,
because
I'm
also
seeing
differences
between
teams.
So
it's
not
a
consistent
pattern.
Okay
and.
A
C
We
we
are
making
some
progress.
I
Emily
is
it's
traveling
at
the
moment,
so
we're
trying
to
clean
up
the
development
dashboard.
Sorry
Becky.
The
short
answer
is
we're
not
we're
not
making
a
ton
of
progress
yet,
but
it's
something
that
I'm
hoping
we
continue
to
build
on.
We
do
have
cycle
time
in
place,
so
that
was
the
new
addition.
Okay,.
A
Sounds
good!
Thank
you.
Moving
on
to
time
to
resolve
customer
bus,
its
training
lowered
as
much
month,
which
is
good
but
I
again,
I,
do
not
see
any
improvements
until
we
roll
out
the
heat
map
and
the
training
program
which
is
scheduled
for
Monday
I'll
touch
on
that
agenda
later
on
in
the
meeting
coming
down
to
this
is
concerning
to
me,
I
haven't
pulled
down
the
list
of
CSV
to
see
what
was
the
bugs
for
this
I.
A
C
C
B
B
I've
spent
maybe
a
day
total
doing
some
research
and
I
record
all
my
notes
here,
still
finding
my
way
around
the
tools
and,
what's
available
part
that
so
the
wording
on
hypothesis
eight
has
to
do
with
the
size
of
the
M,
ours
and
developers
with
less
tenure
at
get
lab,
come
with
a
different
methodology
and
small
I
Mars
compared
to
those
who
are
longer
tenure
so
started
with
the
research
of.
Do
we
even
measure
the
size
of
our
M
are
somewhere?
Can
we
capture
that
information?
B
Do
we
think
it
looks
like
there
was
an
M
R
sent
a
while
back?
It
actually
captures
the
changes
count
and
it's
available
through
the
API,
but
I
didn't
I
couldn't
find
it
anywhere
in
periscope.
So
the
data
is
there
if
I
had
to
write
some
code
to
go
through
every
single
M
R
and
get
the
changes
count
to
be
kind
of
slow.
So
one
of
the
questions
I
have
is:
can
we
get
this
data
somewhere
else
and
that's
listed
down
at
the
bottom
or
somewhere
within
this
doc?
B
That
it's
kind
of
what
I
figured
so
data
is
available,
but
not
easily
retrievable.
At
the
moment,
when
trying
to
correlate
all
the
data
that
I
would
need
for
this
metric
and
show
any
kind
of
graphs
that
would
be
meaningful.
There's
data
available
in
periscope
data
available
and
the
quality
dashboard.
The
team
llamo
will
give
us
the
tenure
thanks
for
neck,
for
pointing
me
to
all
these
data
sources
and
the
changes
count
from
the
API.
B
So
there's
no
one
single
source
where
I
get
all
this
information,
so
I'll
have
to
find
a
way
to
correlate
it
all
together
and
then
I
just
started
writing
down
brainstorming
ideas,
so
I
got
to
combine
all
the
information
together
and
get
running
averages
and
look
for
trends
over
interesting
milestones.
So
I
think
I
think
there
will
be
some
interesting
milestones
at
months,
one
three,
six
and
then
twelve
and
beyond
I
think
it
will
flatten
out.
There
should
be
some
kind
of
plateau
there,
but
I
could
be
wrong.
B
That's
why
you
have
a
hypothesis
and
also
looking
for
size
of
M
Mars
at
the
time.
I
think
the
one
observation
that
I've
seen
as
a
manager
over
the
years
is
typically
new
employees
unless
they
already
know
the
code
very
well,
because
they're
community
contributors,
like
we've,
had
at
puppet
and
we've
had
here
at
get
lab.
They're,
probably
gonna,
take
on
small
and
very
well-defined
issues
just
so
they
can
get
used
to
working
at
the
new
company.
B
The
processes,
the
people,
the
team,
the
code
base
right
so
it'll
be
hard
to
quantify
that
hard
to
quantify
methodologies.
I
think
we're
just
gonna,
see
sizes
and
averages
so
I'm,
not
sure
I'm,
not
sure
how
we
can
quantify
methodologies
in
here.
If
we
can
at
all
so
and
then
the
other
point
I
put
in
there
and
be
interesting
to
count
the
number
of
comments
on
the
developer
M
ours
and
see
if
there
was
a
decline
over
time
as
people
got
used
to
the
code
base
and
had
fewer
questions.
A
B
A
So
it's
neutral,
which
is
good
yeah.
B
That's
the
progress
I've
made
on
eight,
so
I
guess
I
should
put
in
their
next
actions
figure
out
where
I
can
get
that
changes
count
as
part
of
an
ETL
somewhere
in
periscope,
and
probably
do
some
research
see
if
there's
anything
else,
I
don't
have
to
keep
asking
for
incremental
changes
to
an
ETL.
Maybe
they
can
get
all
in
one.
So.
B
A
A
backdrop
for
this
is
the
audit
with
finance
that
we
are
going
through
and
since
it's
going
to
be
in
YouTube,
that's
will
be
on
saying
we
have
a
list
of
projects
that
has
been
imported
there
now,
and
this
will
affect
historical
data
for
good
reasons.
I
think
this
also
solved
the
ask
room
from
you,
Dahlia
and
and
the
secure
team
that
they
were
missing
these
projects.
A
A
D
A
A
A
A
A
Thank
you
for
that.
Moving
on
because
we
are
importing
projects.
I
think
it's
important
to
also
make
sure
that
we
have
Hygiene
in
those
child
projects.
I
think
this
will
affect
Christopher's
exercise
of
labeling
DMR's.
Think
we
have
a
good
hygiene
at
the
big
projects,
see
those
are
like
80%
throughputs,
but
the
edge
cases.
For
example,
a
new
project
is
created,
it's
relatively
new,
nobody
bothered
to
add
a
throughput
label
there
or
a
team
or
a
group
label
there.
A
It
sounds
like
a
grooming,
an
automation,
grooming
template
that
needs
to
be
applied
to
all
these
issues
and
just
have
a
have
a
bot
just
run
through
them.
I
want
to
open
up
for
discussion
here,
because
this
will
affect
the
remaining
20%
that
who
never
get
turned
into
a
colors.
It's
always
going
to
be
great
because
it's
like
some
project
somewhere
that
nobody
was
looking
at
importing
engine.
C
A
I
think
these
are
specific
small
projects
for
a
given
group
yeah
right,
so
we
could
probably
be
like
hey
by
default.
All
in
Mars
in
that
project
always
get
a
group
security
scanning
label
or
pop
station
label.
That's
all
one
part
of
it.
The
other
part
is
through
point
because
we
can't
really
infer
automatically
a
key
one
and
I
know.
Tax
agent
and
feature
has
its
own
nuances.
If
we
can
do
to
improve
their
gotcha.
C
Yeah
I
can
I
can
definitely
pass
this
on
to
the
team.
I
agree
with
you.
We
do
want
to
have
label
hygiene
across
all
em
ours,
like
anything,
you
contribute
make
sure
to
put
a
label.
It
just
I
think
if
we
can
make
some
progress
with
the
the
automation
like
if
we're
able
to
run
the
automation
and
get
the
the
group
label
added
that,
that
would
be
a
great
win,
because
that's
that's
what
I'm
seeing
is
that
reinforcing
and
reiterating
you
know
doing
these
things
manually?
It's
not
been
very
successful.
I
thought
I.
A
C
C
If
you
will
I
I
do
agree
with
you
if
we
can
automate
some
of
the
other
ones,
but
those
are
gonna
get
tricky
because
in
our
effort
to
break
things
into
very
small
Amar's,
there
are
instances
where
you
are
putting
your
first,
mr
and
coming
back
and
adding
a
lot
more
tests
like
the
first,
mr
we'll
have
tests
for
sure,
but
not
all-encompassing.
So
I,
don't
I,
don't
want
to
start
creating
edge.
You
know
edge
logic
and
I'd
much
rather,
the
engineers
define
that
themselves.
Yes,.
A
C
A
C
A
D
C
C
I'll
leave
it
up
to
you
I.
The
other
option
is,
as
we
transition
some
of
this
data
to
periscope
I'll
work
with
Emily
to
see
if
we
can
add
filters,
and
that
would
be
an
easy
way
to
filter
the
mrs
without
a
label
so
that
it's
I
think
right.
Now,
it's
more
of
a
like
a
debugging
exercise,
people
having
a
hard
time
identifying
where
those
missing
ones
are.
Okay,.
A
D
A
Changes
in
the
trash
packages
coming
as
well
so
and
I'm
kind
of
cautioning
the
team
out
to
over
the
chasm,
but
we
do
want
to
go
with
the
time
to
resolve
bugs
and
that's
that's
the
main
actor
there
for
trash
packages.
Okay,
I'm,
moving
on
to
the
next
one
heat
maps
are
coming
in
on
Monday
at
the
nature
of
my
screen,
because
I'm
very
proud
of
the
team.
It's
work
that
went
into
this,
though
there
were
a
few
bucks
on
going
on
the
second
that
prevented
us
from
perfecting
it
this
week.
A
But
essentially
this
is
how
it's
gonna
look
like.
This
is
an
example
for
plan
and
I
feel
bad
picking
plan
all
the
time
with
all
that's
always
the
first
different
scoop
it
comes
up.
Essentially,
a
trash
package
will
not
have
a
heat
map
on
what
the
state
of
curtain
box
for
your
group
looks
like
and
for
plan.
This
is
this
is
a
really
good
metric,
because
we
have
so
much
in
p4
s
force.
A
That's
that's
normal
and
that's
good,
very
small
numbers
that
do
not
have
priority
label,
but
they
do
have
severity
and
and-
and
we
don't
have
three
to
have
no
variance
here
any
which
is
great
so
yeah.
It's
gonna
look
like
this
and
then
the
first
section
is
for
the
product
manager,
the
engineering
manager
and
the
back
of
engineering
manager.
A
Okay.
Next
one
is
on
me
as
well:
I
pinged,
German
slack,
the
only
remaining
customer.
This
one
is
tools
that
my
schedule
get
mostly
managed.
Jeremy
is
been
busy
with
the
release,
of
course,
being
a
release
manager
from
the
product
side
as
well,
so
he's
overbooked.
Vitesse
is
on
it
and
the
next
one
is
about
long
as
well.
So
this
has
migrated
to
the
product
already
and
the
next
iteration,
suggesting
that
we
tried
this
into
time.
Series
data
and
just
have
it
for
per
month.
A
So
they're
gonna
see
one
dimension.
This
is
essentially
how
many
PD
bugs
that
we
missed
the
30-day
mark
after
the
label
v3
that
signify
after
the
quarantine
signified.
The
top
section
is
for
customers
customer
with
bug,
sorry
bug
with
customers,
and
then
the
latter
is
essentially
just
generic
bugs
with
no
question
illegal.
So.
A
Yes,
so
it
would
be
like
three
three
bar
charts
for
even
month.
Like
word,
this
has
been
opened
say
six
months
ago,
whatever
and
as
how
many
P
1
P
2
P
3
the
amount
of
bugs
that
were
beyond
that
threshold.
So
this
would
give
us
like
a
rough
timeline
of
like.
Are
we
boring
down
the
tail
end
or
are
we
putting
down
the
front
end?
Mr.
mark.
C
Yeah
Mike
I
think
the
the
my
proposal
for
for
a
next
iteration
that
would
be
extremely
helpful,
is,
if
we
can,
you
know
either
when
you
click
on
on
this
graph
or
particular
data.
If
it
could
take
you
to
the
query
that
displays
all
the
issues,
because
then
it's
very
actionable
and
we
can
quickly
look
at
this-
highlight
the
the
problems
and
and
easily
transfer
that
into
here's
a
list
that
you
need
to
pay
attention
to
right.
A
I
completely
agree,
that's
on
the
roadmap
and
we
can
appoint
to
the
issue
tracker
with
the
labels.
I
think
I
did
big
dependency
is
curing
by
clearing
by
creation.
Date
is
not
there
and
we
use
milestones
and
labels,
so
some
building
blocks,
who
need
to
be
there
to
have
the
granularity
of
the
time
series
with
the
query.
But
with
this.
C
A
A
Thank
you.
Thanks
for
the
input
on
the
requirements,
I
appreciate
that
going
on
to
the
next
one,
because
12.0
is
a
transition
month.
The
current
mr.
labels
were
added
on
the
seventh
and
I'm,
going
through
the
closed
ones
and
I'm,
probably
just
gonna,
remove
the
miss
deliverable
label,
because
it's
done
it
was
labeled
a
seven,
but
it's
done
now
before
we
close
the
muscle,
so
it's
considered
done.
A
This
is
all
me
this
week,
the
next
one
now
I'm
pushing
for
label
definitions
to
get
merged
by
end
of
the
week
this
week,
I
think
my
last
comment
has
already
best
Olivia
is
concerned,
they're
two
tracks
of
labeling
just
to
set
the
context
here.
Issue
labeling
is
based
on
feature
and
category
labels,
and
that
has
been
work
with
with
product
management
ready
and
the
issues
bubble
up
to
the
direction
pages.
A
What
this
group
a
working
group
is
concerned
about
most
is
throughputs
and
the
auto
labeling
mechanism
would
be
leading
towards
authors
instead
and
then
the
flex,
with
the
flexibility
of
overriding
manually,
so
I
think
the
ad
for
the
twenty
is
a
dress
and
I'm
see
any
reason
why
we
shouldn't
move
over
this
and
I'm
looking
forward
to
to
close
without
any
comments
discussions.
Oh.
A
Okay,
then,
the
last
one
I
have
set
up
a
time
to
do.
The
recording
training
for
tracking
bugs
I
have
invited
Jason,
Dahlia
and
folks
on
in
this
meeting
are
interested
I'll.
Add
you
to
it
as
well,
so
we're
gonna
do
a
run-through
of
the
heat
map,
explaining
what
varieties
and
severe
teas
are
supposed
to
be.
We
clarify
the
severity
guidelines.
A
The
pearly
guidelines
are
very
clear,
so
we
would
start
off
with
that
and
then
possibly
close
up
with
an
AMA
I
like
towards
the
end
of
the
month.
So
after
people
read
this
watch
it
they
can
ask
us
questions
on
hey.
How
can
I
use
severely
and
probably
the
better
schedule
for
my
teams
and
such
since
I'm
good,
okay,
okay,
then
I
would
notice
well.
A
A
Ensure
customer
facing
bucks
have
a
priority
label.
This
is
50%
I,
think
of
it
as
a
female
to
apply.
We
are
charging
the
customer
box
right
now.
We
will
look
at
into
this
later
for
the
rest
of
the
customer
affecting
box
with
no
no,
with
no
priority
and
security.
A
time
to
resolve
as
tooth
is
done.
We
used
to
monitoring
this
I
would
create
this
at
at
50%,
because
we
have
the
metrics
done
already,
but
we
need
to
burn
it
down
or
some
some
number
of
lower.
A
Maybe
maybe
40,
because
the
numbers
up
there
but
yeah
I'll
make
it
him
are
after
this
and
pick
this
up,
40
to
50%
an
effective
iteration
to
the
current
group
trash
packages
to
highlight
customer
defects.
This
is
still
half
way
done.
Heat
maps
are
coming
in
we're
going
to
add
some
context
around
custom
affecting
box
there
later
and
it
will
be
done.
A
B
A
C
We
we
should
I'm
what
I'm
running
through
my
mind,
one
of
the
things
that
I
did,
although
it's
blurry
right
now
is
link
the
video
that
you
Eric
and
I
had
around
throughput.
So
the
throughput
page
should
have
that.
You
know
it
doesn't
look
like
I
finished
the
EMR,
so
I'll
add
an
mr2
link
to
that
video
I
have
been
doing
it
with
all
of
my
managers,
but
I
need
a
recorded
artifact.
So
I'll
get
someone
one
of
the
managers
to
run
through
it
in
a
recorded
form
and
we
can
post
them.
Okay,.