►
From YouTube: CHAOSS Metrics Models Working Group 5/24-5/25/22
Description
Links to minutes from this meeting are on https://chaoss.community/participate.
B
B
C
C
A
C
B
Okay,
um
not
a
problem,
so
just
for
folks
that
aren't
kind
of
following
so
with
community
activity
and
funding.
You
know
one
of
the
things
we're
just
kind
of
talking
about
here
is
structure,
and
I
think
the
next
thing
we
talk
about
is
structure
as
well,
but
we
were
kind
of
following
this
structure
of
algorithm
data,
insight
and
data
for
how
we
present
our
metrics
models
and
then
sean
was
using.
This
is
one
of
the
first
metrics
models,
kind
of
pre-deciding
on
that
structure,
and
I
think
sean
was
gonna
kind
of
wrangle.
D
Okay,
what
is
what
is
the?
uh
What
is
the
relationship
between
the
metrics
model,
markdown
page
and
the
implementation,
this
page
right,
so
community
activity
is
the
metric
model
right,
uh
yeah,
okay,
so
this
is
just
the
implementation
here
right,
yep,
okay,
then
we
have
a.
We
also
have
a
community
activity,
markdown
file,
correct.
D
D
B
E
F
B
F
Sure,
actually,
we
have
a
little
little
bit
discussion
in
the
last
stage
of
this
specific
meeting
about
the
metric
model
report
structure
proposal.
Here
we
realized
that
we
have
a
definition
of
metrics
model
that
kevin
you
mentioned
and
we
have
community
activities
we
have
community.
Welcome,
needs
money,
money,
definition
of
metrics
model
and
also
at
the
same
time,
we
have
the
implementation
of
those
matrix
models.
F
F
F
F
And
also
the
first
two
papers:
actually
they
these
related
to
the
business
ecosystem.
Last
the
first
three
papers
lead
to
the
business
ecosystem
and
our
concept
from
community.
I
mean
open
source
community
ecosystem.
Actually,
I'm
think
of
is
their
original
originated
from
those
business
ecosystem
concept.
F
So
here
for
the
different
item,
I
gave
the
definition
I
have
to
administer.
The
definition
of
the
of
those
items
are
slightly
different
in
the
different
papers,
but
I
could
say
that
they
they
share
the
common
parts
in
the
the
most
of
the
shared
common
part.
So
I
I
gave
the
definition
of
here
so
it
so
from
the
definition.
You
can
see
that
it's
much
more
straightforward
you
can.
You
can
read
directly
to
see
what
it
means.
F
So
it's
part
of
the
I
mean
the
k
areas
of
of
the
whole
ecosystems.
First
first
is
productivity,
the
efficient
with
which
an
ecosystem
converts
input
into
output
and
robustness
and
capability
of
our
ecosystem
to
face
and
survive
directions
and
each
creation
the
ability
to
create
meaningful
diversity
and
survive,
noble
capabilities.
F
G
F
Yeah
and
and
here
code
quality,
of
course,
we
have
definition
and
we're
we
already
have
and
the
implementations
and
these
implementations
like
kevin's
question,
do
we
have
more
than
one
implementations
to
to
for
one
metric
model?
My
answer
is
yes,
and
of
course,
we
could
have
more
than
one
implementations
for
one
single
matrix
model
where
we
can
put
here
so
uh
so.
Under
this
implementation
folder,
we.
G
F
F
F
What
I
mean
the
atomic
metric
definition
is
here
like
the
definition,
the
weight
uh
in
the
in
the
subsection
4.1.
uh
We
have
uh
like
a
contributor
count.
We,
uh
the
definition,
is
how
many
active
committee
authors
will
be
patman
actually
in
past
90
days.
That's
the
way.
I
would
uh
we
add
the
filter
strategy
in
this
metric,
so
we're
clear
to
know
that
how
to
use
this
atomic
magic
in
the
matrix
model,
it's
different
from
the
magic
metric
definition.
F
C
D
F
G
F
F
C
D
C
E
D
D
B
Right
provide
we
provide
that
here,
kevin,
so,
okay,
I
have
the
this.
Is
that
google
doc?
Obviously
you
know
what
I
mean,
so
these
are
the
things
that
would
come
from
the
notebook
and
then
you
can
just
kind
of
read
the
text
you
know
like
here.
It
looks
like
such
and
such
and
such
and
such
is
happening.
Okay,.
F
B
D
Right
so
there
are
uh
obviously
in
the
in
the
the
way
that
these
algorithms
are
put
together.
There
are
some
uh
kind
of
decisions
that
are
being
made
right,
so
you
could
have
so
right
now
we're
just
doing
python
code,
so
the
our
target
user
is
someone
who's
comfortable
with
python
somebody's
comfortable
with
json
right,
other
implementations
could
be
r
or
another
platform.
D
Additionally,
I
think
there
there
are
different
ways
that
you
can
look
at
the
interaction
between
the
metrics
in
these
models
and
different
weights
that
you
can
look
at.
So
I
I
uh
uh
and
I
think
we
had
said
that
uh
multiple
implementations
is
a
possibility
and
I
think
that's
a
good
thing.
I
think
we
need
to
leave
that
open.
B
I
think
we'd
have
to
think
about
I'm
totally
fine
with
that
too.
We'll
have
to
think
about
how
that
would
be
shown
in
this
folder
structure,
because
right
now,
like
the
data,
is
these
two
json
files
which
feed
into
this
one
jupiter
notebook
and
it's
all
really
good,
like
it's
real
world
community
data
that
provides
insights
in
this
notebook.
D
B
B
I
I
Yeah-
and
I
also
um
we
also
mentioned
that
regarding
the
data
folder
and
the
data
insights
folder,
I
think
that
making
the
delineation,
based
upon
the
type
of
data
that
goes
into
the
folder,
doesn't
help
me
as
a
non-coding
contributor
like
at
all,
because
I
want
to
know
when
I
go
into
something
that
I'm
going
to
be
capable
of
understanding.
What's
in
it,.
I
B
I
I
C
D
I
I
also
think
that
if
you
provide
all
three
you've
now
uh
built
an
entire
ecosystem
for
vega
lite,
which
a
lot
of
people
use
uh
very
frequently
uh
within
data
studio
within
power
bi,
it's
basically
like
a
shared
open
source
version
of
all
of
these
data
processors
that
are
already
industry
standard
in
the
first
place.
So
I
think
that
offering
it
in
csv,
r
and
json
is
really
covering
all
of
it.
F
I
I
don't
it's
possible
that
we
can
just
remove
from
a
natural
language
convention
the
type
of
format
because
of
the
extension,
but
if
we
remove
the
type
of
format
file
from
the
extension,
um
we
still
have
to
make
sure
that
the
folder
is
delineating
um
what
the
user
is
expecting
when
they
click
into
that
folder,
and
that,
I
think
that's
the
crux
of
my
main
confusion.
There.
B
B
B
My
my
guess
is
that
we're
gonna
get
a
lot
showing
up
in
json
anyway,
just
to
start
that's
just
because
I'm
looking
at
the
folks
that
are
kind
of
doing
this
model
work
with
you
know
your
team
and
sean,
and
so
we
probably
don't
need
to
like
explicitly
limit
anything
like
say
this
has
to
be
only
json.
My
guess
is
that
that's
the
only
version
that's
just
gonna
show
up
for
a
while.
G
Hi,
I
guess
if
the
data
is
in
the
structured
format,
then
having
json
or
csv
will
serve
the
same
purpose.
But
if
we
have
like
unstructured
embedded
objects
within
a
data
structure,
then
json
is
the
best
option.
So
I
don't.
I
haven't
looked
at
the
data,
so
maybe
those
who
are
doing
it
if
it
is
a
like
proper
structure,
keeping
json
or
csv
will
even
in
the
python
will
work
the
same
way.
B
G
I
see
a
community
activity
as
a
part
of
an
ecosystem
as
a
one
model.
So
am
I
thinking
of
that
as
an
ecosystem
or
I'm
focused
on
the
model
like
I
have
one
model,
how
that
model
is
being
coded
and
implemented
and
defined
are
the
three
things.
So
maybe
the
structure
should
be
in
that
direction,
or
am
I
looking
at
the
robustness
of
that
model?
B
G
D
D
So
uh
I'm
I'm
kind
of
of
the
opinion
that
focus
areas
should
be
more
context
specific
rather
than
something
that
uh
describes
a
measurement
like
productivity
is
kind
of
something
that
you
would
measure.
uh
I
think
for
me.
If
it's,
if
it's
something
that
you
can
measure
there,
then
it
it
kind
of
becomes
uh
difficult
to
grasp
because
it's
uh
it's
kind
of
different
levels
of
abstraction
right.
D
So
uh
productivity
is
something
we
can
measure
and
then
we
we
dig
a
little
deeper
and
we
can
measure
a
specific
thing
about
productivity
and
we
can
dig
a
little
bit
deeper
and
measure
something
even
more
specific
uh
when
you,
when
we
start
when
we
start
getting
in
those
those
different
levels
of
measuring
productivity,
it
becomes
uh
uh
it's
a
little
difficult
to
uh
to
navigate
as
a
focus
area.
So
I
I
would
prefer
that
the
focus
areas
remain
more
descriptive
of
the
context
and
then
the
models
describe
what
we're
measuring.
D
D
I
Honestly,
if
we
were
to
get
rid
of
that
layer,
it
would
honestly
get
rid
of
a
layer
of
abstraction.
For
me,
I
don't
really
feel
like
productivity,
robustness
and
niche
creation
are
really
doing
anything
for
me,
um
but
again
that
also
becomes
very
difficult
to
manage
later
on.
So
the
question
is
how
many
metrics
models
do
we
actually
think
are
going
to
be
necessary
and
is
that
layer
actually
needed
if
we're
not
planning
to
use
serp
like
20
models.
I
I
B
G
So
my
thought
is
like,
for
example,
in
the
productivity.
We
have
code
quality
example
and
we
have
a
definition
in
implementation.
We
have
the
same
definition
in
implementation
for
community
activity
too.
We
have
the
same
definition,
implementation
and
all
these
algorithms
for
other
models
in
other
categories
too.
I
I
F
F
J
Yeah
hi
uh
hi
everyone
uh
I
joined
the
last
asian
pacific
meeting
and
at
that
meeting
I
proposed
that
we
could
categorize
these
metrics
models
into
different
perspective
perspectives.
For
example,
one
model
is
focusing
on
the
artifacts
and
some
other
models
focusing
on
the
people
and
the
community
around
the
around
the
project
or
in
the
ecosystem,
which
I
think
is
can
be
combined
with
this
uh
structure.
D
So
that's
that's
very
similar
to
the
way
the
uh
the
common
working
group
categorizes
their
metrics,
uh
and
I
am
a
I'm
a
fan
of
that
as
well.
I
think
that
uh
uh
it
matches
with
the
the
point
that
I
that
I
was
trying
to
make
earlier,
where
focus
areas
should
be
more
context
defining
rather
than
than
measurement
defining.
So
are
we
looking.
G
D
D
D
B
So
how
could
we
think
about
the
work
that
um
yehui
and
jun
are
bringing
forward
with
kind
of
this
ecosystem
proposal
around
productivity,
robustness
and
niche
creation?
I
under
I
I
understand
what
you're
saying
like
not
a
layer
and
then
a
layer
and
then
a
layer
like
how
can
we
think
about
presenting
what
liang
is
bringing
forward
with
social
and
technical,
along
with
productivity
robustness
like?
How
can
we
narratively
bring
these
together?
J
J
B
I
Yeah
um
what
I
initially
interpreted
lang
to
be
saying
and
uh
kevin's
main
concern
regarding,
uh
if
we're
going
to
be
grouping
metric
groups,
we
need
to
group
them
based
upon
specific
elements
of
culture
within
a
community,
is
how
I
saw
that
being
said.
So
it
kind
of
brought
me
back
to
the
specific
elements
of
culture
in
anthropology.
So
is
this
a
metrics
model
that
pertains
to
the
measurement
of
attitudes
in
a
community?
I
Is
it
the
one
that
pertains
to
values
to
goals,
experiences
practices
in
a
community
or
artifacts,
which
section
of
a
culture
is
a
specific
model
measuring
which
could
easily
be
a
web
graph
that
we
could
place
at
the
top
right
hand,
corner
of
a
model
and
then
using
that
web
graph?
We
can
organize
things
based
upon.
What's
its
primary
measurement.
B
I
I
B
B
G
B
I
So
um
what
you
could
do
like
the
way
that
I
personally
envisioned
it
was,
if
you
look
at
the
definition
of
community
uh
for
anthropology,
it
specifically
has
attitude,
values,
goals,
experiences,
uh
practices
and
artifacts.
So
these
are
the
things
that
you
focus
on
when
you
study
you
can
replace
those
concepts
with
what
you
have
here,
which
is
uh
how
does
how
does
this
metric
um
like?
What
is
the
focus
of
this
metric?
Is
it
more
robustness?
B
A
H
D
The
if
the
goal
is
to
create
a
metrics
model
that
has
utility
for
a
user-
maybe
maybe
we
should
maybe
we
should
actually
survey
some
users
and
uh
see
what
types
of
categories
or
naming
conventions
or
what
makes
the
most
sense
to
them
from
a
user
standpoint
right.
Do
they
wanna
do
they
wanna
have
a
categorization
based
on
these
are
pull
request
models.
A
B
B
D
There
is
a
there
is
an
ongoing
conversation
as
well
about
what,
as
as
I
said
earlier,
what
what
is
a
focus
area,
what
is
a
metrics
model
and
what
is
a
metric,
uh
and
I
think
the
I
think
that
conversation
is
maybe
uh
happening
in
evolution.
Sean
are
you
still
on?
Is
that
the
point
yeah
this
stuff
off?
uh
I
think
that's
where
we
decided.
We
were
going
to
talk
about
that.
So.
B
B
I've
been
kind
of
rebuilding
the
proposed
template
for
metrics
models,
just
kind
of
based
on
what
we've
been
talking
about
here
and
what's
been
showing
up
in
the
json
files.
You
know
I'm
just
trying
to
match
I'm
trying
to
match
the
work
that's
occurring
in
the
template.
So
as
we
build
new
metrics
models,
we
are
following
the
template
so
um
and
then
that's
about
it.