►
Description
Slides: https://chaoss.github.io/website/CHAOSScon/2019EU/slides/Metrics-Company-Led-OSS-Project.pdf
Session: Metrics in a company-led open source project
Ray used community metrics in two very different open source communities over the past 4+ years. One was a foundation hosted project (www.opnfv.org) with dozens of member companies, and the other is a single-company led open source project at GitLab (https://about.gitlab.com/). In this session, Ray will discuss both similarities and different challenges that he has seen when working with metrics in these two communities. In addition, Ray will share learnings during his transition to a company-led project including identification of goals and stakeholders for community metrics. Ray will then discuss how metrics are being used and analyzed at GitLab.
A
All
right
so
I
think
we're
about
five
minutes
behind,
so
I'll
try
to
get
caught
up
so,
like,
like
I,
said
for
those
of
you
who
weren't
here
earlier
in
the
morning
I'm
ray
Pike
up
until
about
seven
months
ago,
I
was
with
the
Linux
Foundation
involved.
A
number
of
the
projects,
in
addition
to
chaos
and
I,
made
a
transition
to
get
lab.
I
started
like
June
25th,
so
so
it
was
kind
of
interesting
when
I
made
a
transition
from
a
foundation
based
project
to
one
company
led
project.
A
The
way
I
started,
looking
at
metrics
obviously
has
changed
so
I
mean,
although
I've
only
been
at
the
company
for
seven
months,
just
went
to
share
the
journey
that
I
went
through
and
sorry
about
the
long
title
for
my
talk.
Obviously
I'm
not
going
to
be
hired
as
an
editor
for
any
new
news
organizations,
but
please
feel
free
to
stop
me.
If
you
have
any
questions.
I
won't
be
offended
a
bit
if
I
don't
go
through
all
of
my
slides.
A
So
so
here's
my
view
of
my
previous
community
up
until
June
22nd,
so
I
started
working
with
folk
folks
from
Victoria.
For
several
years-
and
this
was
sort
of
my
view
of
the
world-
and
what's
sorry,
the
folk
focuses
a
bit
off
I
guess,
but
hopefully
this
slides
are
already
on
the
on
the
on
the
website.
So
you
can
get
a
better
view
there.
Probably,
but
one
of
the
things
that
I
focused
on
was
the
participating
organizations
that
are
contributing
to
a
project
and
they
sorry
I
should
have
mentioned
it's.
The
project
called
OPN
Fe.
A
This
is
in
the
networking
space
in
case
you're,
not
familiar
with
that.
So
this
isn't
even
all
the
organizations
that
were
actively
participating,
but
we
had
about
a
dozen
or
15
that
were
key
participants.
Like
you
know,
companies
like
AT&T,
like
Ericsson,
Huawei
and
others,
so
I
looked
at
a
lot
of
the
contributions
from
an
organizational
perspective
rather
than
like
individual
developers.
I
did
know
a
lot
of
the
key
developers.
Obviously,
but
I
really
focused
on
I
mean
things
like
I
mean
you
probably
heard
the
term
like
the
bus
factor
right.
A
If,
if
this
project
or
foundation
loses
like
X
number
of
participating
organization
like
what's
going
to
happen
to
the
project
and
the
other
thing
that
your
you'll
notice
is
that,
like
a
number
of
repositories,
it
tends
to
be
pretty
large
for
large
collaborative
projects,
I
mean
not
all
of
these
projects
were
active,
I
mean
there
are.
There
are
a
lot
that
I
was
starting
to
like
archive
as
a
community
manager
when
I
was
at
OPM
EFI,
but
you
have
a
lot
of
people.
A
I
mean.
What's
interesting,
though,
was
that
I
mean
it
sorry,
it's
hard
to
read
like
a
number
of
submitters
like
if
you
compare
like
what
we
have
it
get
lab
and
what
we
had
in
OPN
efi.
The
number
is
actually
almost
like
2x
and
I'll
talk
about
this
in
a
minute,
but
a
lot
of
these
are
very
casual
contributors.
A
I
mean
get
lab,
is
a
completely
open
source
and
we
have
a
paid
there's
an
enterprise
version
that
you
pay
for.
Even
that
is
like
open
source
and
we
we
welcome
people's
contributions.
So
we
have
like
users
and
and
people
that
are
passionate
about
ruby.
For
example,
you
know,
submit
merge,
requests
and
then
we,
happily
we
happily
accept
them,
and
that's
one
of
the
things
that's
been
kind
of
cool
about
our
our
product
and
our
our
community.
So
so,
obviously
my
view
has
changed
and
then.
A
And
so
I'll
quickly
talk
about
what's
been
sort
of
different,
you
know
working
in
two
different
communities,
so
the
first
thing
I
noticed
was
the
contributors
motivation
on
why
they're
contributing
to
the
project
for
projects
like
Emil
to
go
as
you're
talking
about,
but
OpenStack
I
mean
there
are
a
number
of
projects
at
the
Learning
Foundation
that
people
are
contributing
to,
but
for
the
most
part,
people
contribute
because
it's
part
of
their
job
I'm.
There
there's
an
expectation
that
you're
spending
X
percentage
of
time
contributing
to
OpenStack.
A
As
an
example,
I
mean
I,
don't
know
how
they
enforce
it,
but
they
there's
a
certain
level
of
expectation
that
you
get
paid
to
work
at
company
X
because
you're
we
want
you
to
contribute
to
these
open-source
projects.
So
when
I
was
when
I
was
working
at
a
foundation,
this
is
like
a
lot
of
motivation
is
professional
right,
it's
part
of
their
job
and
that's
it.
That's
the
reason
why
they
accepted
accepted
the
offer
they
work
for
their
company.
A
So
a
lot
of
it
was
very
professional
and
what
I
noticed
when
I
came
to
get
lab
I
mean
first.
The
first
challenge
I
had
was
I
was
trying
to
figure
out
who
these
people
were
because
I
mean
we
have
hundreds
of
contributors
that
are
contributing
to
get
lab
and
I
had
no
idea
who
they
were
and
their.
Why
they're,
contributing
and
I
had
to
reach
out
to
these
people
on
an
individual
basis.
So
I
felt,
like
I,
was
going
from
a
wholesale
model
to
like
a
retail
model.
A
I
did
like
reach
out
to
these
people
on
an
individual
basis
and
the
reasons
were
very
varied,
but
it
was
closer
to
Hobby
versus
work.
There
are
certainly
a
lot
of
people
that
were
scratching
their
own
itch.
There
are
people
that
are
using
gitlab
for
work,
and
then
there
are
certain
features
that
want
to
tweak
or
new
features
they
wanted
to
introduce.
A
So
there
are
a
certain
number
of
those
people,
but
there
is
also
a
lot
of
people
that
were
very
passionate
about
ruby
and
they
were
looking
for
open
source
project
where
they
they
can
contribute
to,
and
then
one
of
the
things
when
they
did
a
Google
search
was
that
get
lab
kind
of
popped
up,
and
then
this
is
sort
of
their
way
of
giving
back
to
the
Ruby
community,
which
I
thought
was
really
interesting.
So
they
were
literally
doing
this
after
work
like
there.
They
were
there
with
their
day
job.
A
They
come
home
and
then
they're
making
contribution,
whether
it's
making
documentation,
updates
or
actually
fixing
a
bug
or
I
found
every
question.
So
it
was.
It
was
really
eye-opening,
so
these
were
like
almost
like
a
hobbyists
like
they're,
doing
work
and
making
contributions
to
to
get
lab.
So
the
approach
that
I
had
to
take
what
that
was
was
significantly
different
than
it
was
like
a
year
ago.
So
that's
one
difference
and
the
other
other
difference
in
work
style
that
I
had
to
go
through
was
how
I
interact
with
the
community
members.
A
So
when
I
was
at
a
foundation,
open
source
found,
you
know,
base
foundation,
I
met
with
a
lot
of
people
through
meetings
or
like
events
like
this
I
have
developer
events
like
several
times
a
year
and
I
have
a
lot
of
faith.
I
spend
a
lot
of
face
time
with
them
in
a
room.
You
know
on
GoToMeeting
or
zoom,
and
just
working
with
in
real
time
and
what's
now
different
working
at
a
single
company
lead
project.
Is
that
it's
it's?
A
My
interaction
is
like
one-on-one
and
we
have
a
kit
lab
what
we
call
the
core
team.
So
these
are
like
invitation
basis.
If
there
are
people
that
have
been
making
significant
contributions
over
time,
we
invite
them
to
a
group
called
the
core
team
and
you
can
go
to
get
lab
that
come
and
do
a
search
there.
So
we
have
about
like
eight
people
right
now
and
they
made
like
amazing
type
of
contributions,
not
just
in
terms
of
code.
A
There's
one
person
that's
active
because
he
does
amazing
amount
of
like
a
user
support
on
our
IRC
Channel,
like
somebody
posts,
a
question
I'm
trying
to
like
deploy,
get
lab
and
I'm
stuck
here
and
then
Hannes
like
jumps
in
like
unprompted,
like
I
mean
we
have
like
official
support
channel
as
well,
but
he
justs
provide
a
lot
of
community
support.
So
we
have
a
core
group
of
people
that
you
know
a
core
group
of
people
that
we
work
with,
but
their
rest
of
them
that
I
reached
out
to
on
an
individual
basis.
A
I
just
set
up
a
call
like
I
found
this
developer
in
Russia
in
this,
like
middle
of
Russia,
that
I'd,
even
it
wasn't
familiar
with.
But
it's
like
over
the
last
three
years.
He
made
like
a
lot
of
significant
contributions,
but
this
interactions
like
one-on-one
right
there
he's
not
in
a
meeting
with
me,
like
what
twenty
other
people
and
the
other
channel
that
I
really
started.
Using
and
I
was
never
like
active
in
social
media.
I
mean
I.
A
The
only
one
that
I
really
use
on
a
daily
basis
was
Twitter,
but
I
was
basically
a
consumer
of
Twitter
content
that
guy
wasn't
posting
anything
really.
If
you
look
at
before
like
May
of
last
year,
the
only
tweet
that
I
made
was
something
related
to
soccer
like
that
was
it.
It
had
nothing
to
do
at
work,
but
that's
Twitter's,
another
tool
that
I
started
using
a
little
bit
more
actively.
A
I'm,
not
that
I
mean
I,
don't
have
a
whole
lot
of
poster
there
or
anything,
but
that's
the
tool
that
I
started
using
just
so
that
I
can
like
reach
out
to
poor
me
people
more
effectively
and
the
other
communication
mechanism.
That's
gotten
that
that's
very
different
is
my
communication
with
people
is
now
a
more
like.
It's
almost
like
ninety.
First,
five
percent
of
the
time
is
asynchronous.
It's
so
I'm,
not
in
meetings
with
them.
Like
I
mean
a
lot
of
times.
The
developers
are
in
completely
different
time
zone.
A
It
was
even
worse
in
and
what
what
it
was
when
I
was
at
the
Linux
Foundation
but
I.
This
is
something
that
adjustment
that
I
had
to
go
through.
I
have
to
be
more
comfortable
and
willing
to
do
eight
asynchronous
work
with
people,
because
you
know
there
are
all
I
mean
literally
all
over
the
place
and
I
had
to
work
with
these
people
like
on
a
one-on-one
basis
and
and
part
of
it
has
to
do
with
our
company
culture
too,
like
I
mean
we're
get.
A
Lab
is
a
hundred
percent
virtual
organization
and
the
last
count
I
heard
from
HR
was
we
have
people
in
45
different
countries,
so
there
I
can't
sorry
about
that.
I
can't
expect
people
to
be
up
because
I
happen
to
be
out
right
wherever
I
am
so,
the
asynchronous
mode
of
working
I
thought
I
sort
of
had
that
like
well
handle
when
I
was
at
the
Learning
Foundation
I
realized
it
was
still
like
working
with
people
synchronously,
which
isn't
like
a
reasonable
thing
to
do.
A
A
A
A
You
want
to
raise
awareness,
I
mean
you
want
to
convert
them
to
make
a
first
initial
purchase,
and
then
you
want
to
make
them
become
repeat,
buyers
or
repeat
contributors,
so
that
wasn't
very
far
off
I
thought
was
kind
of
odd
question
to
ask
during
the
interview,
but
but
I
try
to
think
about
it
in
those
terms.
Ever
since
ever
since
I
started
working,
so
obviously
things
like
awareness
is
important
and
then
you
probably
saw
in
my
first
slide,
there's
a
there's,
a
tagline.
A
This
is
everyone
can
contribute,
but
that
sort
of
been
our
company
tagline
for
for
several
years
and
we
want
them.
We
want
people
to
understand
that
we
mean
that,
like
when
people
submit
a
merge
requests,
I
mean
merge,
requests
if
you're
not
familiar,
that's
basically
the
same
thing
as
pull
requests
and
get
up
like
if
people
make
a
submission
that
we
want
them
to.
We
want
people
to
understand
that
we're
gonna
treat
those
like
pretty
seriously.
We
don't
like
disregard
those
because
they
came
from
outside
of
get
lab,
but
we
want
people
to
understand
that.
A
That's
you
know.
We
that's
just
that's
more
than
just
a
motto
like
we
want
people
to
believe
it,
so
they
keep
coming
back
for
more
and
then
they
become
part
of
our
regular
community
and
then
the
third
one
is
something
that
you
know.
It's
probably
going
to
be
my
focus
for
the
rest
of
the
year.
Is
we
have
a
lot
of
people
that
are
aware
and
they're
starting
to
contribute,
but
how
do
we
make
them
become
more
regular
contributors
and
how
do
I
do
that?
A
I'm,
not
quite
sure,
but
so
those
are
the
three
things
that
I
pay
attention
to
and
then
I'll
just
talk
about
like
a
difference,
different
metrics
that
I
look
at
on
a
day
to
day
basis,
so
that
I'm,
you
know,
I,
can
keep
track
of
the
work.
I'm
doing
am
I
making
progress
or
what
are
some
of
the
other
things
that
I
need
to
work
on
so
in
terms
of
awareness
or
making,
people
feel
welcome.
A
This
is
something
I
try
to
do
when
I
was
at
the
Linux
Foundation,
but
there
wasn't
a
really
good
way
of
doing
it,
like
in
Garrity's,
were
using
Garrard
in
no
pmf
in
community.
How
do
I
like
recognize
like
first-time
contributors
or
make
him
feel
welcome,
and
then
we
have
a
couple
of
tools
that
we
we
use
for
that
I
mean
there's
this
little
little
badge
up
there.
Next
to
the
merge
request
like
it,
it
highlights
that
this
is
persons
like
first
contribution
and
then
I
also
added
a
label.
A
Basically,
do
a
query
you
once
a
month
and
say
for
our
last
release:
I
mean
our
current
releases.
11.8,
that's
going
to
be
released
on
February,
22nd
I,
just
do
acquired
and
say
like
we're
the
first
time
contributors
for
that
release
and
then
it
this
makes
it
easier
for
me
to
reach
out
to
them
and
and
then
thank
him
and
appreciate
them
for
their
work,
and
this
is
something
that
I
really
did
when
I
was
in
OPN.
A
Fbi
I
thought
about
doing
this,
but
I
didn't
have
to
really
prioritize
it
because
not
I
mean
they
didn't
have
just
me
to
work
with.
But
if
they're,
if
they
came
from
like
Huawei,
is
an
example,
they
have
other
coaches
within
the
company
that
can
help
them
out.
So
that
wasn't
something
that
I
really
focused
on.
But
this
is
something
that's
pretty
important
like
on
a
day-to-day
basis.
I
get
lab,
so
I
could
like
reach
out
to
them
and
then
make
sure
that
I
know
who
these
people
are,
and
they
can
come
back
for
more.
A
The
other
one
is:
are
we
being
responsive
to
our
community
members?
So
the
contribution
that
comes
in
through
a
merge
request?
If
we
looked
at
the
identity
or
the
person,
and
if
we
know
that
that's
not
a
get
lab
employee,
it
gets
flagged
as
a
community
contribution.
It's
just
automatically
as
a
label.
A
It's
pretty
frequent,
so
one
of
the
things
I
asked
our
merger
quest,
coaches
or
maintainer
to
do
is
make
sure
that
a
label
is
added
which
is
called
like
milestone
now,
I
I
had
I
wanted
to
like
use
this
image,
because
I
thought
it
was
like
really
funny.
So
there's
we
have
an
engineer
named
fatty
and
like
he
and
I,
had
a
couple
of
back-and-forth
because
he
he
would
like
to
forget
to
like
a
put
a
label
on
on
a
on
a
patch
that
has
been
merged.
So
this
is
like
my
third
interaction
with
Fatih.
A
Just
on
the
milestone,
I
said
dude,
you
forgot,
they
had
the
label
and
then
he
basically
apologized,
and
then
he
just
threw
that
pick.
I
thought
it
was
like
pretty
hilarious.
I
was
I
was
worried.
That
I
was
being
just
too
heavy-handed
with
people
on
taking
care
of
community
contribution,
but
just
showed
that
fatty
has
a
sense
of
humor
and
then
here's
another
report
that
I
looked
at
I
look
at
so
it's
a
medium
time
to
open
like
is
it.
Are
things
getting
merged
and
reviewed
in
a
reasonable
time
frame?
A
So
I
did
a
comparison
between
community
contributions
and
internal
development
work.
Obviously
it's
like
the
internal
work
was
a
lot
faster
and
I
think
there
are
a
lot
of
reasons
for
that.
Like
I
mean
our
developers
that
if
they've
been
working
to
get
left
for
a
while,
they
understand
the
architecture
they
don't
really
have
to
like
rework
their
their
patches
and
community
contribution.
Obviously
it's
it's
reasonable
think
that
it's
going
to
take
a
little
longer.
A
I
just
want
to
make
sure
that
differences
for
reasonable
and
then,
when
I
look
at
the
average
like
less
than
four
days.
That's
not
bad
I
mean.
Could
we
do
better
sure,
and
then
there
are
certain
merge
requests
that
doesn't
get
merged
in
a
reasonable
timeframe,
then
that
we
still
need
to
work
on,
but
something
that
I
tried
to
keep
in
and
keep
an
eye
out
for
and
make
sure
for
each
release
that
there
isn't
a
whole
lot
of
fluctuation
or,
if
there's
a
fluctuation.
Hopefully
it's
for
the
better.
A
You
know:
I'll
try
to
wrap
things
up,
really
quick,
I'm,
I,
guess
still
five
minutes
over
so-
and
this
is
the
last
one
is,
you
know,
I'm
trying
to
still
figure
out
what
I
need
to
do:
I'm,
converting
casual
contributors
to
regular
contributors,
and
then
this
is
an
article
from
last
year,
I
believe
on
open
source.
Comm
but
the
author
basically
talks
about
there
are
like
a
five
or
six
important
factors
that
helps
you
convert
casual
contributors
through
more
regular
contributors.
I
thought
it
was
like
a
really
good
article.
A
I
mean
things
like
like
the
cultural
norm
of
the
organization
is
important,
how
they
feel
welcomed,
do
they
identify
with
the
rest
of
the
community
members,
etc,
etc.
So
how
that
gets
translated
into
my
day-to-day
job
and
get
lab
I'm
still
need
to
figure.
It's
figure
this
out.
But
my
motivation
comes
from
this
data.
Point
I
actually
shared
this
with
our
core
team
members
like
last
month.
A
Emphasis,
Bonnie
I,
believe
the
name
like
that
contributions
like
less
than
20
over
the
last
five
years
right.
So
you
know
because
what
I
told
the
core
team
members
who
said
this
band
here
I'd
like
this
people
to
be
more
like
fifty
people
versus
like
fifteen
or
twenty
that
we
have
today.
So
these
are
the
regular
contributors
that
have
been
doing
more
than
like
five
merger
quest
merge
per
year,
because
I
pulled
that
number
out
of
the
out
of
the
thin
there.
A
A
I,
don't
know
what
the
right
target
is,
but
what
I
like
to
do
is
develop
a
deep
bench
so
that
there's
a
reliable
number
of
contributions
that
are
coming
in
from
outside
of
the
company
I
mean
the
main
motivation
for
this,
and
this
is
something
that
I
got
questions
on
from
the
core
team
members.
It's
not
just
to
get
just
more
code
from
outside
of
gitlab,
but
their
perspective
is
very
important
because
they
are
thinking
you
very
differently
a
lot
of
times
from
their
users
perspective,
not
like
internal
engineer,
that's
getting
paid
by
gitlab.
A
So
it's
not
just
like
a
free
technical
work
that
that
I
appreciate
it's.
The
outside
perspective
is
I,
think
it's
just
as
valuable.
So
the
one
of
the
things
that
I
I
stress
was
that
I
like
to
get
that
deeper
pool
of
people,
making
contributions
and
bringing
different
perspectives,
and
so
that's
you
know
how
I'm
going
to
do
that
this
year,
I
I,
don't
really
know
if
you
have
any
suggestions
or
ideas
on
it,
I'd
love
to
hear
more,
but
any
like
one
or
two
quick
questions
before
I
turn
things
over
to
the
next
speaker.
A
I
mean
our
like,
so
the
funny
thing
about
guilt
lab
like
we
try
to
be
transparent
about
everything
and
then
the
problem
with
that
is
is
just
way
too
much.
Information
like
I
can't
even
digest
like
everything
that
gets
published,
but
if
you
go
to
our
product
page
sort
of
that,
what
we
want
to
do
for
the
next
few
releases,
that's
informations
all
public
and
even
like
our
retrospective,
and
when
we
do
a
kickoff
for
the
next
launch
that
all
those
calls
are
public,
so
I
mean
whether
people
know
where
to
find
them.
A
That's
a
different
story,
but
so
the
the
plan
or
roadmap
for
our
product
is
is
open.
But
in
terms
of
like
your
original
question,
I
think
there
are
a
couple
of
different
categories.
One
is
like
scratching
your
own
itch
like
there
are
certain
features,
I
didn't
like
they
wanted
to
improve
it
once
that's
done,
they
they're
done
like
they
don't
sometimes
like
a
lot
of
times.
A
They
don't
come
back
or
there
are
other
things,
though,
fixes
that
were
like
really
simple,
like
there
was
a
good
documentation
error,
and
then
we
love,
like
you,
know
like
ill
to
go.
Saying
this.
We
don't
think
of
is
something
that's
trivial
like
this
helping
QAR
work
right,
so
people
make
like
quick
fixes
like
documentation,
fixes
and
then
a
lot
of
times
they
don't
like,
for
whatever
reason
they
don't
come
back
so
I
think
it's
it
sort
of
varies.
But
you
know
we
want
to
make
sure
that
the
contributing
experience
is
good.