►
From YouTube: Development Metrics Working Group 2019-08-08
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
B
So
on
the
mrs
merch
per
month
we
basically
did
have
our
highest
months
so
far,
the
including
country
community
country
she
was
at
20
97.
We
were
just
three
mr
shorter
2,100
that
converted
to
basically
our
goal
is
to
increased
by
about
20
percent
throughput
peak-to-peak
between
three
months,
so
we
hit
85
percent
of
it.
If
you
would
have
done,
we
did
1791
previous
one.
So
it's
like
basically
360
additional,
mr,
so
we
were
shooting
for
him.
We
hit
about
300
those
360.
So
that's
do.
B
A
B
Even
notice
that
that's
that's
it's
true,
that's
a
it'll
be
interesting
to
see.
We
have
August
September
and
October.
August
actually
has
22
working
day,
so
we
may
actually
see
August
be
the
higher
I
runner.
Those
September
and
October
obviously
have
additional
people
ramping
and
none
of
those
fall
into
either
contribute
or
a
holiday
month.
So
it'll
be
interesting
to
see
how
that
was
kind
of
three
months
by
out,
so
we
may
see
John
progression,
but
I
would
expect
it,
maybe
a
little
bit
lighter
towards
the
end
of
the
quarter.
B
Hiring
and
essentially
average
I
Mars
per
author,
so
we've
kind
of
acknowledged
that
we're
not
quite
back
to
ten
but
we're
wanting
to
keep
in
that
8
8
to
10
range,
and
we
definitely
have
did
that
in
the
past
quarter
or
contribute
withstanding.
The
one
aspect
is
I
would
love
to
see
a
nine
in
there,
but
we're
just
we're
not
we're
not
trending
as
well
as
we
would
hope
there.
At
this
point.
B
Need
to
air
kind
of
talked
about
this.
He
had
an
additional
item
for
things.
We
need
to
try
I,
don't
think
that
would
close
out
the
working
group,
though
I
think
that'd
be
more
of
a
development
initiative.
So
I
don't
think
that's
that's
a
good
call
that
I
need
to
I
still
need
to
circle
back
with
him
and
then,
since
I
was
out
yesterday,
I
really
didn't
have
a
chance
to
follow
up
on.
On
that
specific
one.
A
Okay,
thank
you
moving
on
to
the
last
one,
so
we
were
still
monitoring.
This
I
did
change
the
exit
criteria
not
to
cover
aggressively
or
lowering
the
number
cuz
I,
don't
think
it
will
go
lower.
A
Josh
made
me
speak
more
to
this,
but
we
have
made
some
improvements
and
adding
more
metrics
to
to
help
teams,
groom
their
backlog
better
and
also
focusing
on
P
1
and
P
2,
so
that
should
help
in
the
interim.
A
D
All
right,
so,
on
hypothesis,
eight,
the
hypothesis
was
that
new
engineers
come
in
with
a
different
methodology
where
they
have
smaller
and
Mars
are
dreamers,
geez
backward
sorry
come
in
different
methodology
compared
to
those
longer
tenured,
more
productive,
so
yeah,
the
hypothesis
I'm
saying
the
new
engineers
come
in
to
smaller
in
Mars
and
actually
the
trends
that
I'm
seeing
in
the
data
proved
the
opposite
or
show
the
opposite.
So
this
chart
on
the
left
is
showing
engineers
who
have
less
than
a
year's
worth
of
tenure,
and
actually
there
M
our
size
and
the
only
metric.
D
D
So,
what's
the
trend
over
time
of
everybody-
and
this
is
actually
the
trend
of
all
the
data
over
time
and
there's
no
there's-
no
trend,
it's
kind
of
up
and
down
right
on
the
average
number
of
files
affected
by
EMR
inclusive
of
everything
in
that
CSV
that
I
downloaded
so
and
the
number
across
the
bottom
is
just
the
spread
of
between
created
and
merged
or
I'm.
Sorry,
tenure
yeah.
D
D
D
B
A
C
B
That's
essentially,
what
we
want
is
the
DIF
summary
if
you
do
like
a
git
diff
summary
it'll,
give
you
like
the
number
of
lines
of
lines,
changed
and
I.
Think
that
would
be
a
good
proxy,
like
I
feel
like
number
of
files
changes,
not
necessarily
good
proxy,
but
lines
changed
as
a
products
good
proxy.
It's
not
necessarily
lines
of
code,
but
I've.
You
know,
let's
not,
let's,
let's
just
do
a
first
order
or
some
a
first
order
with
that.
B
That's
what
we're
looking
for-
and
you
know,
there's
two
ways
you
can
potentially
do
that.
One
way
is:
is
you
get
all
the
commits
that
are
associated
with
it,
and
then
you
go
figure
out
what
the
changes
are
associated
with
it.
Then
that's
how
you
count
it
other
ways
as
if
the
m
are
actually
had
that
data
in
it.
B
A
D
B
D
C
B
B
D
So
the
where
we
left
it
off
laughs
was
when
Christopher
was
last
with
us
looking
at
this
graph,
and
so
we
go
back
the
peaks
where
we
had
the
highest
averages
were
this
one,
so
June
2018
and
January
2018.
If
we
look
at
the
velocity
church
on
throughput,
those
were
the
two
peaks
that
we
had
and
we
overlaid
it
with
our
growth
in
hiring
around
that
same
time,
so
we
were
relatively
flat.
D
Yes,
yeah,
so
I
pull
an
interim
team
me
I'm.
All
looked
at
the
did
a
lot
of
data
cleanup
on
titles
and
made
sure
to
exclude
anyone
from
the
CSV
download
the
author
name.
That
was
not
in
T
mammal,
so
yellow
only
includes
folks
that
were
in
the
team,
yellow
file
engineers,
sorry
engineers
and
then
employees
is
everyone.
That's
in
enamel,
and
then
all
this
includes
everybody
that
was
in
the
CSV
download
its.
D
B
Alright,
the
your
labels,
good
one
thing
that
we're
talking
about
is,
is
whether
or
not
we
actually
do
it
specifically
to
the
development
department
mmm-hmm.
So
because
of
that
we've
been
using
the
term
developers,
but
your
labels
correct
here
I
would
have
asked
for
it
to
change
if
it
was
based
on
the
development
departments,
working
yeah.
B
D
B
A
You
have
only
one
left
cool,
let's
jump
on
TV
at
the
rest
of
the
agenda,
we've
added
new
charts
for
the
requests
from
last
time.
This
is
the
deliverable
delivered
deliverables
versus
other
issues,
and
this
would
also
allow
Joshua
to
to
add
this
data
directly
to
his
spreadsheet
as
well.
So
let
me
just
share
my
screen,
so
I
can
show
you
that.
A
Yes,
I
did
sunspot
check
like
it
is.
It
is
actually
correct
if
we
look
at
the
initial
tracker,
it's
43,
because
somebody
added
one
manually
last
night
so
that
or
after
this
thing
happened,
but
this
is
P
43
tomorrow
and
it's
the
same
as
what
the
tracker
says.
If
you
look
for
Miserables,
miss
12.0
label.
B
Right
and
we're
in
the
middle
of
12
to
right
we're
in
a
middle
talk
to
ya,
so
this
session
I
had
a
discussion
about
this
yesterday.
Your
feels
like
two
months
ago.
Basically,
you
know
the
ratio
of
deliverables
to
undefined
close
issues
is
just
unbelievably
high.
This
one
right,
yeah
and
the
question
would
be
is-
is
whether
or
not
we
should
be
going
back
and
having
the
team
like
it
just
feels
like
if.
B
25
percent,
the
75
percent
of
the
work
is
undefined.
It
just
feels
like
we're
not
anywhere
even
close
to
matching
up
with
expectations
here,
at
least
that's
kind
of
my
take
on
it
from
that
perspective,
so
we're
trying
to
figure
out,
we
go
back.
Cuz
like
it's
just
did
find
a
couple
teams
that
weren't
labeling.
So
that's
that's
the
first
problems,
hygiene
and
you
know-
should
we
be
actually
be
asking
teams
to
go
back
and
mark
things
deliverables
if
they
were
intended
for
that
release
and
just
never
got
marked
correctly.
A
E
I
think
I
don't
have
it
I,
don't
have
a
good
theory
on,
unlike
on
the
non
Global's,
and
that
volume
I
actually
don't
know
something
to
be
curious
to
double-click
I.
Think
obviously,
part
of
the
contributing
factor
here
is
that
we
are
labeling
it
reliably.
So
we
need
to
figure
that
out,
but
not
sure
why
that
would
happen
unless
we're
not
going
in
and
labeling
them
like
midstream
right
so
like
if
a
bug
comes
in,
for
example,
after
planning
happened,
we're
not
we're
not
going
in
and
then
labeling
it
a
little
more
so
that's
possible.
E
I
much
prefer
use,
deliverable
and
then
have
a
single
prioritized
list,
but
some
folks
have
blue
versus
stretch
still
so
I
think
we
need
Center
as
a
process
like
that.
I
think
will
go
a
long
way
to
help
make
that
tart
useful,
because
right
now
not
everyone
uses
it.
The
hygiene
problems
are
sort
of
all
over
the
map.
So
I
will
do
that.
As
far
as
the
planning
process
that'll
get
that
done
this
week,
I
mentioned
hopefully
work
well
and
then
we'll
hopefully
them
ask
for
the
teams.
E
E
I
send
that,
along
at
least
as
like
as
I
find
them,
I
thrashed
me
off
in
a
stretch.
Words
I,
think
that
then
they
had
some
really
critical
issues
that
they
wanted
to
try
and
deliver
like
three
of
them,
and
so
they
made
those
the
only
deliver
bolts,
the
entire
release,
entire
milestone
and
all
the
rest
of
the
stuff
wasn't
stretch
I'm
just
to
make
it.
You
know
super
clear
that
these
three
were.
You
know
the
most
important
things
so.
E
A
A
A
A
A
B
Yeah
I
guess
what
I'm,
what
I'm
trying
to
reconcile
in
my
head?
Is
they
mark
me?
This
is
something
to
think
about.
You
were
asking
about
what
the
label
is.
If,
if
it's
undefined
I
could
see,
if
it's
a
bug
that
we
list
out
the
bugs
it
bugs
the
bugs
are
marked
deliverable,
then
they
go
and
deliverable
if
they're,
not
if
bugs,
are
marked
undefined
or
not
mark
deliverable,
then
we
mark
them.
We
have
that
category.
Does
that
make
sense?
B
So,
like
the
rules,
the
most
important
thing,
those
are
the
things
are
committed
to,
or
you
know
that
we
said
this
is
above
the
cut
line
we
were
doing
planning
bugs
is
things
that
we
fixed,
that
you
know
just
yeah
stuff
happens
and
we
prioritize
that
and
then
there's
still
this
category
of
undefined
and
what
I'm
I'm
trying
to
reconcile
in
my
head.
Is
this
like
right
now?
B
The
way
it
looks
at
it
is
like
if,
even
if
with
the
200
bugs
on
top
of
that,
it's
still
like
50%
of
our
work
is
undefined
and
like
what
Josh
is
seeing
is?
Is
that,
like
you
know,
for
the
things
that
we
supposedly
project
in
our
calls
is
50%
we're
all
getting
50%
of
those
targets?
So
if
something
is
just
not
making
sense
to
me,
because
it's
not
about
the
cut
line,
it's
it's
not
above
the
cut
line
where
we
prioritize
it.
B
Yet
it's
so
important
that
you
know
50%
of
our
work
is,
is
undefined
right,
like
not
even
not
even
talked
about.
This
is
kind
of
the
way
I'm
bit
that
could
very
well
be
very
important
things
that
engineering
needs
to
be
doing
but,
like
it
still
feels
like
like
100
percent
purga
zation
is
not
happening.
That
would
be
my
assertion
at
this
point
based
on
this
50
percent
and
I'm
just
kind
of
scratching
my
head
still
on.
What's
going
on
so.
D
New
guy
question:
was
there
a
definition
somewhere
in
the
handbook
on
what
is
a
deliverable
versus
what
is
not
I'm?
Looking
at
the
memory
team
and
the
deliverable
label,
this
is
really
the
first
time
I've
heard
of
this
I.
Might
it
might
be
my
fault,
but
most
of
ours
are
not
labeled
as
deliverable
are
charged
pretty
awful.
E
D
So
then,
thank
you
for
that
clarification,
I'll
start
paying
more
attention
to
that.
What
about
things
that
are
discovered
in
the
middle
of
the
milestone
so
like
we
start
working
on
an
issue
and
we
discover
we
need
to
do
two
more
issues
immediately
within
that
milestone.
Do
we
then
mark
those
as
well
as
deliverables,
even
though
they
weren't
planned
prior
to
the
kickoff
or
the
start
of
the
milestone,
yeah.
E
That's
I
think
the
question
we
need
to
answer.
The
same
thing
goes
for
like
you
know
you
ship
flow
to
and
you
find
three
bugs
that
are
pretty
high
priority
and
you
just
jump
on
them
yeah
and
and
should
those
should
you
just
and
and
then
you
know,
you're
gonna,
you're,
gonna
work
on
them
and
you're.
Then
gonna
go
push
out.
Potentially
some
things
you
did
plan
on
roll
a
release,
so
I
think
the
question
is:
how
do
we?
You
know
what
I'm
gonna
do
that
situation?
E
Do
you
want
to
try
and
track
like
how
much
of
what
we
planned
makes
it
through
or
is
it
more
of
a
plan
which
is
unplanned
well,
I?
Guess,
I?
Guess
if
it's
just
plan
for
some
plan,
you
still
want
to
then
try
and
track
some
of
that
stuff
but
like
as
in
what
originated
in
but
I
feel
like.
We
need
some
provisions,
some
good
process
to
deal
with
these
things
that
come
up
right
like
a
bug
or
whatnot
I.
D
Was
originally
viewing
these
charts,
as
the
Green
was
planned
and
gray
was
unplanned,
but
I
think
there's
a
combination
of
not
knowing.
We
need
to
label
these
as
deliverable
and
I.
Think
if
we
want
to
do
the
plan
versus
unplanned,
we
need
to
be
explicit,
otherwise
we're
probably
muddying
the
waters
there.
So
yeah.
E
I
think
because
back
I
think
what
ultimately,
what
ultimate
metrics
going
to
get
from
this
I
think
I.
Think
the
the
workflow
around
the
cut
line
is
is
helpful
and
useful
right.
So,
as
a
p.m.
I
have
a
prioritized
list,
the
engineering
manager
can
just
go
through
and
say
Bing
Bing,
Bing,
Bing
Bing
at
me.
Stop
here,
right,
I,
think
that
that
works
out
really
well
and
then
the
question
that
is
for
another
use
of
it,
which
is
planners
unplanned.
E
How
should
we
handle
these
midstream
changes
right
because
we
PM's
also
don't
want
to
be
stuck
in
that
situation,
where
you
know
we
find
something
else
more
important
to
work
on.
Even
if
it's
not
a
bug
and
we
don't
feel
like
we
can
shift
it
because
we're
gonna
be
duffed
a
fine
plan
work.
So
we
should
try
and
say
agile,
as
you
possibly
can,
but
so
I
think
that
that's
the
question
around
sort
of
the
plan
versus
I'm
behind
word.
Okay,.
A
B
E
A
Sorry
I'm
taking
the
last
one
so
a
bit
on
labor
migration.
Eighty
percent
of
the
data
should
be
correct,
already,
there's
to
some
edge
cases
where
there
are
multiple
team
labels
that
our
team
will
work
through,
but
it
should
be
kosher
now
and
answer
all
our
questions.
I
will
review
the
exit
criteria
next
time
around
I
think
we're
making
may
be
making
a
lot
of
progress.
So
let's
talk
about
that
next
time,
interest
at
the
time
and
thank
you
Mark
courage
for
taking
the
investigative
work
on
and
to
work
on
label
migration
appreciate.