►
From YouTube: CHAOSS Metrics Models Working Group March 28 2023
Description
Minutes from this meeting can be found here: https://docs.google.com/document/d/1qE2duE73NbEOHzRRAbqgPSUSAfdH-AmvLruOu7uCIYg/edit
A
So
one
of
the
things
that
I
I
did
want
to
bring
up.
So
it's
good
to
see
everybody
by
the
way.
One
of
the
things
that
I
did
want
to
bring
up
today
was
something
that
we've
been
talking
about
in
the
ospo
community,
so
there
seems
to
be
at
least
from
my
perspective,
a
real
push
from
ospo's
to
start
articulating
value,
not
just
from
the
talk
we
had
a
week
ago,
but
from
like
social
channels
from
conferences.
There's
there's
really.
B
A
A
need
here,
dot
I
saw
you
nodding,
I'm
guessing.
C
A
A
D
A
A
But
there
are
a
couple
of
large
organizations
that
have
been
making
some
changes,
whether
it's
like
full,
so
I
think
this
question
is
probably
something
we
should
address
or
try
to
help
move
forward.
I
think
there
would
be
a
lot
of
really
good
reception
in
the
ospo
community.
A
A
C
I
think
I
think
that's
an
important
distinction
too,
because
I
think
it
really
is
like
people
say
Roi,
but
they
don't
actually
mean
the
financial
definition
of
Roi.
They
really
just
mean,
like
the
showing
value,
showing
a
return
for
the
number
of
people
that
they
have
have
working
on
it,
so
I
think
I
don't
want
anybody
to
get
hung
up
on
Roi
as
a
financial
measure,
because
I
think
it
really
is
about
showing
the
value
that
the
people
in
that
hospital
bring
to
the
company.
A
Fair
and
then
the
other
thing
that
and
I
think
that
that's
completely
fair
and.
D
A
Think
the
other
thing
that
well
two
points,
one
done:
I
think
in
that
conversation
with
respect
to
moving
it
off
the
financial
component.
You
had
done
a
nice
articulation
of
trying
to
tie
it
to
strategy.
A
Was
this
this
per
through
here,
so
I
think
this
will
give
us
some
good
places
to
begin
and
then
the
other
conversation
that
came
up.
Do
you
remember
this
one
Don?
It
was
ospo
work
versus
open
source
work
and
so
just
kind
of
differentiating
between
those
as
well
and
I.
Think
the
the
focus
was
to
concentrate
on
ospo
work
first
and
then
that
would
help
us
understand
how
we're
working
in
open
source
communities
so.
C
A
A
C
I
think
that's
the
place
to
start
I
think
that,
ideally,
you
go
beyond
that,
but
I
think
especially
right
now,
given
the
financial
climate
that
we're
in
I
think
being
able
to
justify
that
you're
also
is,
is
showing
some
sort
of
value
to
the
company.
I
think
is,
is
what
a
lot
of
people
are
focused
on
right
now:
okay,.
A
Right
on
super
helpful,
thank
you.
Okay,
so
I
don't
know
really
how
to
begin
this
here,
like
so
I'm
hoping
we
could
talk
about
how
to
even
begin
to
produce
metrics
models
that
might
be
helpful
for
ospos.
B
I
I
I,
I
I,
now
saw
it
that
Ry
such
quickly
pop
up
for
from
the
discussion
you
know
from
from
chaos.
Also
because
recently
you
know
we
just
hold
up
also
submit
in
last
weekend
and
to
get
in
China
and-
and
you
know
that
there
are
a
lot
of
people
from
different
hospitals
and
they
came
together
to
discuss
around
the
things
and
the.
The
final
question
is
that
how
much
value
Hospital
products
are
introduced
to
your
company
and
they
do
have
strong
reason
to
remain
these
people.
B
B
So
there
are
quite
a
lot
of
questions
and,
and
we
discuss
what
what
things
can
can
really
also
do
for
for
one
company.
B
Actually,
the
people
who
has
been
working,
who
have
been
working
for
the
company
I
mean
the
hospital
for
for
a
long
long
time
size
they
say,
have
a
couple
things
to
do
first
and
they
should
produce
a
the
whole
procedure
of
precise,
like
how
to
manage
the
open
sources,
the
company
use
and
how
to
choose
that.
B
And
if
some
company
want,
if
some
open
source
you
know
initialized
form
from
company
and
how
to
open
source
thing
and
to
you
know,
based
on
from
the,
in
the
point
of
view,
important
of
compliance,
security
and
and
the
risk
and
some
other
risk
and
and
better
there's
final
question
always
have
been
questioned
by
by
high
level
people
in
company
lights
and
what
kind
of
relationship
between
the
something
like
Roi
and
that's
why
the
produced
by
by
open
source
or
also
so
that's
that
question
you
know,
there's
no
final
answer
still
from
that
meeting
and,
of
course,
from
not
from
us
and
you
know,
but
as
we
discussed
today,
I
have
some
thinkings
about
that.
B
If
we
have
to
mention
a
if,
if
we
just
narrow
down
the
discussion
topic
just
for
the
Open
Source
initialized
by
company
and
what
kind
of
outline
returns
from
this
open
source,
so
in
my
my
thinking
that
we
have
to
fund
a
director
connection
between
the
business
value
and
and
something
very
cleared,
the
fund
in
in
the
company
to
for
my
thinking
is
that
the
usage
count,
adoption
or
adoption
count
and
the
so-called
adoption
count
means
how
many
users
has
been
used.
B
You
are
so
open
source
software,
and,
and
would
they
love
to
use
that
or
they
have
to
use
that
with
without
any
other
choice.
So
anyway,
adoptions
is
to
is
some
Metric
to
have
very
close
connections
with
the
business
value
and
based
on
this
adoption
count.
C
And
I
think
that's
a
particularly
challenging
thing
to
measure
right
is
is
adoption
because
you
you
just
don't
you
don't
know
who's
adopted
your
software
because
they
may
have,
they
may
have
downloaded
the
release
from
GitHub.
They
may
have
downloaded
it
from
a
package
manager
from
an
operating
system.
C
There's
so
many
different
ways
to
get
software
that
it
measuring
adoption
is,
is
really
is
really
hard.
I
mean
that's
that's
what
everybody
wants
to
measure
for
sure
and
I
think
we
need
to
find
I
think
we
need
to
find
more
creative
ways
which
I
think
this
maybe
was
what
you
were
kind
of
getting
at,
but
I
think
we
need
to
find
creative
ways
to
use
other
metrics.
C
A
C
Think
at
the
start,
but
that's
a
contribution,
usually
a
contribution
metric
as
opposed
to
an
adoption
metric,
but
it
does
get
at
adoption
because
those
organizations
wouldn't
be
putting
people
on
the
project
if,
if
they
weren't
using
it
in
in
some
way
so
I
think
it's
it
is.
It
is
one
proxy.
You
know.
Overall
contributors
is
kind
of
another
proxy
for
that,
because
the
more
people
you
have
contributing
to
it,
they're
probably
interested
in
using
it
in
some
way.
But
it's
not
a
not
a
direct
measure.
It's
more
again.
Another
proxy.
B
Yeah
I
still
want
a
response
to
Dawn's
question
how
to
mirror
adoption
count.
It's
you
know
you
recently
opened
Ai
and
the
track
GDP
gave
me
a
lot
of
thinking
around
it,
so
it
currently
we
only
got.
We
only
ask
if
we
want
to
get
some
value
of
metric.
We
are
thinking
for
the
trace
data
existed
in
the
GitHub
or
some
other
code
host
platform
like
JT
and,
of
course,
some
other
data
sources
exist,
provide
some
rest
API,
but
it's
kind
of
it
looks
like
we.
B
We
lack
of
some
mirror
way
to
measure
the
the
you
know.
The
evaluations
from
the
people
who
use
that
who
use
those
softwares
and
this
evaluation
could
come,
could
come
from
a
forum
like
a
discussion
all
came
from
a
book
or
some
books.
They
would
recommend
you
to
use
some
softwares
on
some
books.
Like
some
technical
books,
they
will
introduce
several
choices
around
some
tech
around
some
technique
and
they
give
you
the
writer's
author
would
give
you
some
choice.
They
give
they
would
evaluate
and
also
some
papers.
B
B
So
if
we
can
use
such
data
sources
to
help
us
to
evaluate
or
to
predict
to
filter
adoptions
of
some
softwares
or
open
source
softwares,
it
would
be
it
would
be
really
helpful
and
thinking
and
and
I'm
still
thinking
that-
and
you
know
as
and
on
your
right,
because
for
some
business
softwares
it
it's
kind
of
a
business
product.
We
can
easily
know
who
are
using
our
product
as
a
software
or
some
other
tools,
because
they
would
buy
or
buy
a
license
or
some
something
like
that,
but
for
the
open
source.
B
It's
really
hard
to
do
so
because
I'm
thinking
like
as
very
beginning
of
Open
Source
and
no
one
care
actually
who
are
using
their
the
open
source.
You
know
developed
or
or
initialized
by
the
people
like
Richard
stoma.
They
want
to
share
the
free
spirit
of
the
open
source.
They
don't
care
who
use
that.
B
But
you
know
in
the
recent
10
years
many
Commercial
Business
collaborations
corporations
joined
open
source.
These
words,
they
are
thinking
for
the
business
value,
of
course,
so
that's
making
the
adoptions
or
usage
and
quite
important
to
help
them
to
evaluate
how
much
value
such
open
source
software
has
been
produced
by
for
their
company
so
thinking,
maybe
the
value
on
this
open
source
has
been
changed
a
lot
in
recent
10
years,
so
we
have
to
thinking
for
some
ways
to
produce
a
way
to
measure
this
adoption.
B
C
Yeah-
and
you
know
you
mentioned
kind
of
natural
natural
language
Discovery
and
that
that
got
me
thinking,
I
think
that's
I,
think
that's
a
really
good
idea
and
I
think
we
can
build
on
that
in
other
ways,
so
I
spent
a
year
or
two
doing
kind
of
market
research
and
competitive
analysis
and
one
of
the
ways
that
I
would
figure
out
what
our
competitors
was
doing
was
I
looked
at
their
job
postings
and
you
can
look
at
things
like
job
postings
and
tell
if
a
company's
using
a
technology,
because
they're
looking
to
hire
people
who
know
about
a
particular
technology,
so
job
postings
are
another
good
one.
C
B
C
Before
now
that
we
have
more
sophisticated
kind
of
AI
and
machine
learning
sort
of
models,
we
can
get
at
this
I
think
in
more
more
sophisticated
ways
than
what
we
were
able
to
yeah.
You
know
10
years
ago.
B
So
so,
I'm
thinking,
if
people
from
I
mean
the
evaluation
from
the
different.
You
know
nature
language
channels
like
job
posting
blogs.
B
If
they
some
sometimes
would
happen.
Okay
the
the
the
not
that
positive
evaluations
came
came
up
from
those
Niche
language
channels,
they
would
say:
okay,
there
must
be.
Some
problem
happened
in
this
open
source.
C
Yeah
yeah
I
think
that
I
think
I
think
sentiment
is
really
important
for
sure
whether
it's
positive
or
negative.
A
Believe
it
or
not,
we
have
I
think
it
would
be
one
year.
A
So
it's
angled
a
little
bit
different
I
think
it's
about
how
I,
as
a
person
can
understand
where
I
should
be
positioned.
If
I
recall,
when
we
were,
you
know
like
what
should
I
be
working
on
like
if
I'm
in
a
class
I'm
like
I,
tell
my
students,
you
should
take
a
class
on
kubernetes.
You
know,
because
it's
going
to
help
you
down
the
road
kind
of
thing,
so
I
think
that
was
kind
of
kind
of
the
nature
of
this
originally.
But.
B
B
D
A
D
B
Yeah,
so
so
I'm
thinking,
if
we
couldn't-
and
you
know,
discussed
either
pop
up
some
new
ideas,
what
kind
of
nature
language
channels
we
can
utilize
for
these
options-
and
you
know
I
I-
think
we
can.
We
can
have
a
I
mean
the
strong
enough
indicators
for
us
to
predict
adoptions
of
of
one
open
source
software,
and
we
can
we.
We
can
verify
that
actually
I'm
thinking
to
for
some
specific
technical
areas.
B
I,
hopefully
you
know
all
there
are
always
some
areas.
They
would
open
up
their
adoption
count.
So
we
can
use
that
for
some.
You
know
like
train
data
and
we
can
use
test
that.
A
Okay,
so
this
is
super
helpful
I'm,
looking
at
any
other
metrics
in
here,
just.
D
Three
to
the
start
of
the
predictor
work
about
to
predict
some
something.
D
B
B
D
D
Yeah,
maybe
some
magic
can
be
tagged
right,
but
can
be
attacked
can
be
attacked
with
predict,
but
some
of
the
metrics
can
it's
just.
The
result
cannot
predict
right.
A
A
C
I
also
added
I,
just
put
it
in
the
notes
in
the
kind
of
list
of
analysis,
but
we
could
look
at
LinkedIn
and
other
social,
like
user
profile
functions
and
technology,
which
indicates
somebody
working
at
that
company
is
working
on
it.
True.
B
A
A
Least,
that's
for
sure.
Is
there
anything
else
so
again,
a
lot
of
these
are
these
metrics
that
we
have
up
here
are
kind
of
from
the
old
value
working
group
uh-huh.
A
So
is
there
any
and
we
had
the
one
job
opportunities
here
in
row.
17.
was
there
anything
else
in
here
that.
B
Stands
out,
it's
like
a
projector
recommendability.
It's
that's!
The
evaluation
directly
from
from
the
users,
I
think
all
developers,
I.
B
B
A
To
you
and
you
who
we
talk,
this
seemed
to
resonate
with
you
as
a
potential
past
forward.
Are
there
like
thinking
about
the
conversation
that
ospo
would
need
to
bring
to
management.
B
C
C
C
If
you
go
back
to
that
list,
there
was
another
one
that
I
thought
might
be
organizational
project
skill
demand,
maybe
because,
if
they're
looking
for
a
particular
skill
in
an
open
source
technology,
then
they
probably
are
using
that
technology
yeah,
so
that
that
would
also
be
a
proxy
for
use.
I
think.
A
C
C
I
would
just
break
that
out.
I
would
do
it
as
a
separate
section
in
the
and
just
list
some
of
the
metrics.
That
might
be
that
we
can
look
at.
A
Right
something
like
that:
what
are
the
other
ones,
organizational
project,
skill
demand,
what.
C
C
Okay,
job
opportunities.
A
C
Those
are
the
three
that
look
most
most
relevant.
C
B
Yeah
I
think
we
we
make
with
concrete.
Maybe
maybe
we
can
centralize
the
into
one
one,
my
trick
to
say
the
nature,
language,
Channel
or
or
something
like
that
or
talk
so
yeah
I
mean
I.
I
would
like
to
use
some
Metric
to
say
our
indicator
metric
to
tell
me
how
much
how
much
connections
or
collaboration.
B
B
D
A
C
I
do
yes,
absolutely
I
will
also
miss
that
one.
So
if
you
can,
that
I
was
going
to
mention
that
to
you
in
slack,
if
you
could
run
that
one
that
would
be
good.
A
B
And
would
you
would
you
pop
up
the
radio
record
well,
I
mean
the
real
meeting
record
of
the
auspost
next
time,
I
mean
if
yeah,
if
edit
base
also
missed
that
meeting.
A
B
A
We're
that's
great.
We
had
a
couple
others
that
I
just
wanted
to
mention
the
starter
project.
Health
metrics
model
is
finally
complete,
so
that
is
now.
C
I
think
we
still
have
some
confusion
around
how
to
implement
the
new
ratio.
One.
B
Exactly
especially
for
that
matrix
by
Imagine,
called
yeah,
yeah
change
requires
the
closure,
racial
and
I
pop
up
some
some
examples
of
cases.
B
B
C
A
C
B
Yeah
I
gave
I
gave
some
examples.
You
know
to
to
be
understand
the
to
help
understand
this
metric
definition
and.
B
I
mean
there's
one
example:
if
we
follow
the
current
definition
of
this
crr
is
that
Community
a
if
they
that's
I,
must
have
forgot
that.
Thank
you.
C
So
so
in
in
general.
Well,
while
you
look
at
that
again,
the
way
the
way
I
look
at
this
is
I,
especially
for
something
like
a
change
request:
closure
ratio,
I
I,
try
not
to
compare
multiple
communities
against
each
other,
but
instead
I
try
to
look
for
Trends
in
the
ratio
to
indicate
Improvement
or
decline
in
Project
health.
So
that's
that's
typically
how
how
I
would
how
I
would
use
it
so
I'm,
looking
for
I'm
looking
for
projects
that
you
know
their
change,
request
ratio
is
getting
higher
or
lower.
C
B
Would
use
it
the
reason
I
want
to
compare
two
communities,
especially
when
I
come
to
some
specific
area.
I,
don't
really
familiar
with
these
areas.
I
want
to
compare
those
two
open
source
software
which
covers
some
close
or
similar
technical
Focus.
B
B
And,
of
course,
like
you
mentioned,
if
you
already
are
very
familiar
with
this
software,
you
just
care
about
itself.
That's
that's
that
doesn't
matter
so.
The
question
is
that,
should
we
have
to
considering
the
broadcast
from
that
that
haven't
been
processed
from
the
very
beginning,
or
we
only
considering
the
proquest
in
some
period?
B
So
maybe
both
we
are
thinking.
My
thinking,
I
think
they
are
all
how
they
are
specific
requirements.
Are
there
are
meaningful
posts?
B
How
about?
We
have
a
two
definition
or
of
data
display
for
this
metric
to
tell
people
in
the
in
the
past
90
days
and,
for
example,
and
in
the
past
90
days
we
have
a
100
pull
requests
created
and
we
already
closed
our
precise
detail
thing,
so
this
ratio
would
be
the
50
count,
50
percent
and
and
also
we
considering
the
pull
requests
created
before
the
latest
United
States,
better
and-
and
we
would
also
count
the
proquest
processed.
B
And
in
evading,
within
the
in
current
in
the
latest
90
days,
that's
the
definition
from
from
current
definition
of
the
of
this
metric.
D
C
And
then
the
reason
I
mean
I,
think
I
think
that's
probably
an
implementation
detail
that
you
can
decide.
I
mean
that's,
that's
sort
of
you're
applying
an
extra
filter
on
the
metric.
So
maybe
maybe
we
need
to
better
Define
the
filters
on
the
individual
metric
definition,
because
I
think
that
what
you're
doing
is
really
you're
filtering
out
some
of
the
really
old
on
pull
requests
that
are
open
because
you
really
care
about
whether
or
not
they're
closing
the
pull
requests
that
come
in
now.
C
C
That's
true
yeah,
so
what,
from
my
perspective
like
I
I,
don't
want
that
extra
filter
like
I
want
for
all
time,
because
I
want
people
to
go
back
and
close
those
crusty
old,
pull
requests
that
are
never
ever
going
to
be
merged,
but
but
I
think
you
know,
I
mean
I.
Think
your
use
case
is
is
totally
valid.
I
think
that's
a
filter
on
the
metric.
Can
you
click
on
the
metric
itself
and
let's
just
have
a
look
at
what
the
what
we've
put
in
the
filters.
Yeah.
D
C
C
Yeah
so
I
think
you're
applying
this
first
filter
to
your
so
you're
you're,
showing
that,
like
the,
if
you
show
the
change
request,
closure
ratio
for
for
all
time,
but
I
think
you're
also
having
a
like
a
second
one.
That's
filtered
by
date
ranges
so
just
like
the
past
you're,
just
looking
at
the
pull
requests
that
have
come
in
over
a
period
of
time.
D
C
A
B
A
Okay,
okay,
great
good,
okay,
great
so
I
guess,
maybe
just
the
last
thing.
C
D
A
This
we'll
get
this
blog
post
out,
but
this
is
done
so
our
summary
from
chaoscon
Europe
is,
is
completed,
so
feel
free
to
take
a
look
at
this
I
think
there's
some
really
great
stuff
in
here
I
actually
loved
how
cha
looking
at
this,
how
chaoscon
went
and
I'm
actually
really
excited
about
chaos
Con
in
Vancouver,
because
I
think
it's
gonna
to
result
in
really
similar
things
and
it'll
help
us
really
focus
on
on
what
we're
working
on
and
reinforce
what
we're
working
on
too.
A
B
B
D
A
D
A
All
right,
so
we
are
at
the
end
of
time.
Thank
you,
everybody
for
a
really
great
conversation
and
I
think
we
helped
to
start
some
of
the
value,
conversation
and
I
think
we're
finalizing
the
the
metric
model
that
Don
you
had
put
forward
so
Yahoo
thanks
for
your
thoughts
on
that
as
well.