►
From YouTube: CDF TOC 07.02.19 - Screwdriver.cd project presentation
Description
Screwdriver.cd project team presentation to the Continuous Delivery Foundation Technical Oversight Committee
A
B
Great
well,
listen!
Thank
you!
Thank
you
for
that,
and
really
thank
you
to
the
foundation
for
giving
us
the
time
and
the
opportunity
to
present
what
you
know
what
we
built
and
what
we
think
is
gonna
be
really
interesting
for
the
open
source
community
at
large.
My
name
is
Gil
yahuda
and
what
I
really
wanted
to
do
was
introduce
the
team
and
just
kick
off
what
what
we'd
like
to
do
here.
B
So
on
the
call
with
me,
we
have
vineeth
and
jitan
from
the
screwdriver
team
and
we
also
have
ashley
Rosalie
and
myself
from
the
open
source
team.
We're
at
Verizon
Media,
but
we're
also
from
Yahoo
and
Verizon
media
as
you'll
see
in
the
presentation
is,
is
a
company
that
was
born
out
of
a
merger
of
Yahoo,
AOL
and
a
bunch
of
other
companies.
So
we
have
really
a
yahoo
story
and
it's
evolution.
B
We
built
a
really
interesting,
CI
CD
solution
that
we
use
across
the
company.
We
built
it
for
the
open
source
community.
We
built
it
because
we
needed
it,
but
we
also
think
that
other
other
folks
would
need
it
and
would
be
able
to
use
it,
and
it
would
be
appropriate
in
in
this
neutral
home,
for
CI
CD
or
for
CD
technology.
B
So
we
wanted
to
go
through
a
little
bit
of
the
history
of
what
we
built
and
why-
and
we
also
wanted
to
share
details
as
to
you
know
what
it
does
and
how
it
does
it
and
then
feel
questions
from
from
the
foundation
about
it
and
really
the
objective
is
to
see
that
it
becomes
accepted
into
the
foundation
through
the
Foundation's
process,
I
guess
as
an
incubator
project
and
eventually
three
classless
as
a
as
a
top-level.
However,
the
foundation
works
with
respect
to
those
phone
to
to
the
product.
C
I'm
trying
to
share
my
screen,
it
says
I
cannot
spot
screen
share,
while
other
participant
is
sharing
all
right.
Let
me
stop
sign
about
that.
All
right.
Can
you
see
my
screen
there?
We
go
all
right.
So,
first
of
all,
thank
you.
Everyone
for
giving
us
the
opportunity
to
come
here
and
present
about
Scott
ever
I'm
tiffany
manuel
and
the
product
owner
for
screwdriver
within
verizon
media.
C
C
So
big
thanks
for
that,
and
thanks
for
the
opportunity
for
presenting
here
and
about
who
we
are
so
Verizon
media
is
home
to
a
lot
of
products
like
AOL,
Huffington,
Post,
Yahoo,
Finance
Mail,
which
are
used
by
over
a
billion
users
and
scoot
over
CDs.
This
came
out
of
Yahoo
and
this
is
the
build
tool
which
is
powering
the
build
test
and
deploy
life
cycles
for
the
entire
product
at
Verizon
media
all
right
and
our
engineers.
There
build
products
across
different
applications
from
both
iOS
Android
applications,
server-side
applications
and
in
2012.
C
C
Pods
could
ever
our
users
come
to
screwdriver
and
take
simply
come
and
say:
oh,
this
is
my
build
configuration
which
they
can
write
in
symbols
jung-hwan,
and
we
have
this
notion
of
pipeline
a
zip
code
which
our
users
will
provide,
and
they
check
that
into
your
sauce
management
system,
whether
that
is
with
love
or
get
her
pocket
up
in
the
price.
The
scooter
will
understands
that
and
handles
and
creates
your
build
pipelines
on
your
own.
So
it
worked
for
us
for
a
growl.
C
Initially,
scooter
was
just
an
Orchestrator
on
top
of
Jenkins
be
provided,
and
why,
where
our
users
can
come
and
say,
okay
used
use.
This
configuration
to
build
my
application,
whether
that's
web
application
or
a
mobile
application.
But
at
some
point
we
hit
a
wall
in
terms
of
the
code
which
we
can
provide
users.
They
wandered
a
lot
of
cool
workflow
features,
running
Bulls
in
parallel
running
it
across
latest
containers.
At
that
point,
the
containers
were
getting
a
doctorate
in
the
industry.
C
So
that's
when
we
decided
to
redesign
screwdriver
from
ground
up
based
on
the
tremendous
developer
feedback
we
received
from
our
our
users.
So
when
we
design
the
score
ever
before
that
that's
the
one
which
we
offer
sourced,
we
designed
it
in
a
way
that
it's
not
just
specific
for
the
needs
of
one
two
three
we
want.
C
The
park,
any
ties
to
any
culture
technologies
we
use
all
right
and,
as
a
result,
we
designed
and
develops
code
ever
dot
CD,
which
we
open
sourced
in
2016
right
and
so
when
we,
when
we
designs
could
have
a
dot,
sedum
were
getting
getting.
Students
abilities
was
coming
into
picture,
so
we
totally
imagined
a
build
system
where
the
bills
are
actually
short-lived
containers
and
at
the
same
time,
we
didn't
want
to
make
sure
that
we
are
not
tied
specifically
to
containers
and
whatever
system
we
come
up
with.
C
It
has
to
work
for
Radek
use
cases,
because
there
are
people
who
are
using
Jenkins
and
there
are
people
who
are
you've
seen
your
kinetics,
and
we
designed
it
in
a
way
that
the
system
has
the
flexibility
to
use
whatever
the
trust
administration
determines,
that
their
users
needs
so
on.
A
very
high
level
of
this
is
how
the
SCADA,
our
business
process
flows,
the
user
comes,
and
they
creates
a
project
through
a
screwdriver
or
using
your
SEM.
It
goes
to
our
API
and
our
API
picks
up
an
execution
engine.
C
This
execution
engine
is
what
ultimately
runs
your
bills
and
that
kinda
Jenkins
or
that
can
be
kubernetes
or
that
we
just
pure
docker
and
once
the
build
is
completed.
It
archives,
the
data
using
log
storage
using
cash,
and
it
publishes
a
bunch
of
artifacts
to
the
artifact
store
of
your
organization's
choice
and.
C
In
a
nutshell,
the
design
philosophy
of
screwdriver
just
ask
your
build.
You
publish
your
artifacts
when
you
deploy
that
your
infrastructure.
So
as
far
as
the
developer
is
concerned,
it's
all
about
you
commit
your
code,
you
test
your
code,
you
release
it
and
it
goes
on
like
a
cycle
and
towards
the
left
side
of
the
current
slide,
you
can
see
all
it
takes
for
creating
a
screwdriver
pipeline.
As
far
as
the
user
is
concerned,
you
simply
write
a
number:
what
are
your
jobs
and
the
relationship
between
them
and
behind
the
scenes?
C
The
SKA
driver
will
pick
what
executor
it's
configured
birth
and
that
can
be
changed.
I
can
be
talker
or
that
can
be
Copernicus,
and
it
runs
in
that
specific
executor,
and
it's
also
designed
in
mind
with
a
way
that
you
can
pick
and
choose
what
components
you
are
in
she
needs
or
they
can
be
Posterous
or
it
can
be
Manya
DB
or
my
squirrel
or
world
what
application
can
run.
C
You
can
run
your
application
at
scale
in
kubernetes
or
all
the
segments
on
the
mast,
plane,
plane,
docker
containers,
cuke
overview
on
stats,
so
within
Verizon
media
Bilan
over
50,000,
daily
bills,
and
we
have
over
45,000
projects
and
we
support
over
10
programming
languages
and
a
lot
of
platforms
which
of
our
users
comes
and
creates.
These
platforms
are
where
the
scaling
happens.
This
is
where
a
team
comes
and
say
this
is
my
CD
solution,
and
these
are
the
tools.
C
C
I
can
just
go
ahead
and
we
can
take
all
the
questions.
Start.
Okay,
okay,
thank
you.
So
so
this
is
what
I
briefly
mentioned
earlier,
that
all
the
bills
are
containerized
and
at
the
same
time
you
have
the
option
to
pick
different
executors
and
they
also
have
designed
an
executor
based
on
open
souls,
hyper
container
technology
which
enables
used
to
run
a
container
as
a
VM.
C
So
we
support
Fanning
fan
out,
use
cases
easy
or
or
bike
or
flows
connecting
on
pipelines
using
remote
triggers
and
if
you
want
to
run,
word
builds
periodically
more
jobs
and
if
you
want
to
establish
build
constraints
like
please
windows
like
don't,
run
the
build
during
this
period
or
of
the
jobs
using
constraints.
The
workflow
infrastructure
of
screwdriver
enables
this
I
just
want
to
quickly
show
you
build
pipeline
which
follows.
C
It
follows
in
standard
so
right
now
working
and
my
screen
the
UI
for
screwdriver
to
the
right
side.
You
will
see
these
are
all
the
events
which
are
basically
commits
happening
on,
or
you
will
get
instance
and
on
the
left
side
is
a
single
run
of
the
screwdriver
pipeline.
So
this
current
pipeline
shows
that
this
this
particular
event
started
off
as
an
external
trigger
it
started
the
build
job,
then
it
found
out
to
run
two
other
jobs.
Give.
D
B
B
And
while
you're
doing
that,
jitan
I'll
just
remind
and
you
anyone
there's
a
chat
in
zoom.
So
if
anyone
has
a
question
that
once
the
end
wants
to
queue
up,
the
question
please
feel
free
to
use
that
chat.
Will
we're
watching
it
so
that
we
could
ensure
that
you
know
that
everyone's
questions
are
addressed.
Yeah
and.
C
Then
you
can
go
right
now.
Yes,
okay,
all
right!
So
sorry
about
that.
To
my
right
side,
you
will
see
all
the
events
which
are
basically
a
history
of
ripples
and
on
the
left
side
you
will
see
a
single
run
of
this
could
have
a
pipeline,
so
screwdriver
looks
at
this
file
calls
whatever
Toriyama
and
whatever
user
has
found
forged
in
the
Yama's
could
ever
build
this
UI
out
of
it
on
in
this
case,
you
are
anyway
left
side.
You
can
see
that
this
pipeline
started
off
with
a
bill
job.
C
Then
it
fine
to
do
separate
jobs
called
deploying
east
and
west,
and
once
this
two
jobs
are
completed,
that
is
whether
engine
will
fan
in
one
single
job
called
the
test.
So
this
is
a
very
basic
flow
which
contains
a
bunch
of
serial
and
parallel
bills,
and
if
you
have
any
pull
requests
opens,
then
they
will
will
come
here
right
now.
I
don't
have
any
and
you
saw
how
they
start
the
pipeline.
C
C
C
All
right,
so,
can
you
see
in
a
presentation,
yep,
okay,
back
to
the
presentation
in
terms
of
workflow
capabilities,
we
also
support
advanced
workflow
is
like
if
you
have
a
moon
irako
and
if
you
use
branches
for
your
development
or
if
you
want
to
keep
us
whatever
conflicts
upright,
we
call
that
external
conflict
and
another
workflow
feature
we
provide
is
pulled
across
workflow.
So
we
had
a
lot
of
use
cases
where
users
wants
to
run
the
ender
pipeline
even
on
a
pull
request.
Even
before
a
pull
request
gets
merged.
C
They
want
to
see
how
that
changes
behave
so
that
you
can
deploy
that
too.
It
just
infrastructure
run
your
test
cases
against
it,
bring
up
an
application,
so
politico's
workflow
enables
all
those
features
so
lot
of
this,
or
almost
all
these
features
were
designed
based
on
the
tremendous
developer
feedback
we
caught
from
our
internal
users,
and
then
we
designed
it.
We
made
sure
that
it's
in
a
way
that
is
extensible
and
any
develop
organize
a
developers
in
organizations
they
can
find
this,
which
is
useful
and
use
it
in
build
infrastructure.
C
Also,
we
provide
a
native
build
cache
feature
where
users
can
say:
ok,
cache
this
file
or
cache
this
directory,
and
with
this
they
can
move
bits
of
data
between
jobs
and
so
that,
if
it's
not
yet
ready
to
be
persisted
into
an
artifact
storage
and
they
can
use
that
it
can
also
be
used
for
boosting
up
your
performance
by
caching.
Your
most
frequently
used
dependencies
like
maven
downloads
or
no
jeaious.
What
you'll
notice
modules
also
we
we
have
tied
up
with
Sona
cube
for
giving
static
bill
code
analysis.
C
So
screwdriver
has
this
notion
of
micro
services
within
our
infrastructure,
and
so
my
cube
is
just
one
of
them
where
users
can
send
their
build
data
for
offline
analysis,
so
they
can
do
code
smells
they
can
get
coverage
information
and
they
can
know
the
number
of
tests
which
were
run
as
possible
and
also
see
how
we
have
this
feature
called
build
metrics.
But
then
you
can
get
detailed
insights
into
how
you
are
pipeline
behaves.
So
let
me
show
you
quickly
on
how
bill
metrics
is
going
to
look
okay.
Let
me
how
should
I
write
kPa?
C
Okay,
so
this
is.
This
is
a
pipeline
which
actually
builds
screwdriver
itself,
and
this
is
the
bill
metrics
for
the
pipeline
or
a
period
of
month.
So,
as
you
can
see,
we
show
information
on
the
total
duration.
A
build
event
is
taking
and
overall
time
time
it
takes.
You
can
do
deeper
into
that
event,
seeing
what
are
the
build
times
for
specific
jobs.
C
Also,
if
you
scroll
down,
you
can
even
dig
deeper
into
that
specific
job.
Let's
say:
I
select
the
broad
job
and
look
at
how
much
time
each
individual
steps
within
the
jobs
cook.
So
in
this
case,
as
I
can
see,
the
majority
of
the
time
is
taken
by
our
test.
Job
and
I
can
click
on
it
to
deep
link
into
the
specific
just
step,
to
see
the
logs
for
the
step
and
see
how
much
time
the
step
took
off.
So
let
me
go
back
to
my
presentation
that
was
about
Bill
metrics.
B
C
Yeah,
yes,
yes,
I.
Think
one
of
the
specific
question
about
consolidating
I'm
gonna
answer
that
in
just
a
second
with
the
next
slide,
all
I.
Can
you
see
my
presentation
now
yeah?
Okay?
So
so
there
was
a
question
on
how
can
we
make
sure
that
we
don't
repeat
the
steps
again
and
again?
So
this
is
how
we
do
it
with
me
how
this
feature
is
called
templates
and
commands.
This
is
designed
in
a
way
that
the
screwdriver
is
opened
up
for,
subject
matter
experts,
so
the
best
practices
around
how
you
do
a
pill.
C
You
can
encapsulate
that
into
a
template
and
typically
see
what
a
template
is.
Okay,
so
I'll
switch
to
my
computer
in
a
while,
but
I
just
want
to
talk
about
what
templates
and
commands
are
a
bit
more.
So
this
opens
up.
This
could
have
a
platform
or
subject
matter
expert
like
if
you
are
say
you
are
in
node.js
expert
and
you
can
try
out
the
direction
in
your
organization.
How
nodejs
bills
happens
that
you
can
define
steps.
It
says
this
is
how
you
build
it.
C
This
is
how
we
test
it,
and
this
is
how
you
deploy
it.
It
should
have
certain
code
coverage
or
it
should
meet
certain
quality
constraints.
You
can
build
all
those
intelligence
into
a
template
and
any
user
of
who
is
building
no
js'.
You
can
say
that
you
have
to
use
all
you
can
use
this
template
as
an
option
and
once
they
use
this
all
this
encapsulated
things
that
take
bigger
it
for
free
all,
right
and
I'm.
C
Thinking
goes
for
share
commands,
so
shared
commands,
so
template
is
for
a
job
and
command
is
for
a
stuff,
so
you
can
say
okay,
so
how
do
you
want
to
use
your
space?
Your
tool?
Let's
say
I,
want
to
use
spinnaker
for
doing
my
deployments.
You
can
build
a
shared
command
out
of
that
and
expose
that
in
a
way
that
any
user
in
screwdriver
can
use,
can
use
a
shell
command
and
that
command
can
be
for
doing
deployments.
C
E
C
Okay,
so
this
is
salt
templates
look
like
so
in
this
case
they
are
named
spaced
and
let's
say
these
are
the
templates
under
Python
namespace?
Alright,
so
all
these
templates,
so
templates
can
do
one
specific
thing.
Let's
say
this
is
a
template
for
Python
valdek
type,
and
this
is
how
users
will
use
it.
They
will
simply
say:
template,
Python,
well
dead
type,
now
behind
the
scenes.
C
What
this
template
does
is
it
can
use
two
containers,
one
is
Ubuntu
or
something
cat
mini
Linux
and
by
default
it
is
as
the
mini
Linux,
and
these
are
all
the
steps
which
are
encapsulated
with
the
template.
So
the
template
owner,
who
in
this
case
is
an
expert
in
Python
they
they
they
define
the
template
and
only
think
the
users
needs
to
do
is
just
prefer
the
template
in
their
job.
C
So
now
this
means
of
the
run
whatever
steps
which
are
defined
in
this
template,
and
this
is
how
you
can
use
you-
can
avoid
solving
the
same
problem
again
and
again,
create
a
template
and
create
a
template
and
let
the
users
in
your
organization
start
using
it.
So
they
get
all
the
shared
steps
for
free
all
right.
So
now,
switching
back
to
my
presentation.
C
So
that
was
support,
templates
and
commands,
and
so
so
this
is
what
the
temperature
and
commands
basically
allows
their
developers
to
integrate
the
CD
toolings
of
your
choice.
So
using
this
you
can
bring
in
tools
like
Tecton
or
spinnaker
within
your
infrastructure
and
so
to
speak
about
couple
of
integrations.
We
have
within
Verizon
media,
so
we
have
a
lot
of
subject
matter:
expert
cells
which
we
use
Hulu's
ansible,
and
to
be
how
also
tooling
switch
integrate
with
services
like
sauce,
labs
and
pasa.
C
C
Yeah,
so
speaking
of
external
contribution,
we
collaborate
a
lot
with
young
Japan.
They
are
one
of
our
core
core
contributors,
so
some
of
the
features
like
shared
commands
like
that,
was
entirely
designed
and
implemented
by
Japan
and
we
would
be
glue
bi-weekly
sink
ups,
where
the
priorities
of
both
teams
are
brought
up
and
we
decide.
Ok.
This
is
our
key
for
this
week
and
I
work
on
that.
So
they
work
pretty
much
independently
from
us
and
enjoy
the
future
development.
C
And
these
are
couple
of
things
we
have
sorry,
so
we
have
our
blog
and
if
any
free
new
feature
v2,
we
publish
a
blog
post
based
on
how
it's
done
on
how
it
stopped
the
usage
or
useful
guide
our
code
of
contact
and
if
you
are
interested
in
contributing
information,
how
that
can
happen
and
if
you
cover
view
on
our
decision
making
process.
So
our
priorities
are
established.
C
We
have
an
a
public,
github
Kanban
board
got
any
issues
which
are
which
how
users
create
business
where
it
get
prioritized
and
right
now
the
priorities
are
determined
based
on
the
requirements
of
the
two
primary
companies
which
are
involved
in
the
work,
which
is
Verizon,
media
and
Yahoo
Japan
and
all
the
design
decisions
they
are
pretty
well
documented
and
within
within
public
github
repositories,
and
for
release
methodologies,
use
magnetic
police
to
track
package
versions
and
all
the
components
of
screwdriver.
We
have
three
main
components:
UI
API
store.
C
C
B
B
Let's
see
we
go
through
that
yeah,
let
me
let
me
let
me
see
if
I
can
frame
them
for
you.
There
were
three
questions
that
came
in
so
first
of
all,
thank
you.
Thank
you
for
sharing
that
jitan.
During
the
course
of
the
presentation,
I
detected
three
questions
or
three
themes
that
came
up
regularly
on
some
of
the
more
addressed,
so
let
me
frame
them
and
then
we
can
see
if
there
are
other
questions,
one
of
them
had
to
do
with
external
collaborators.
C
Sure,
thanks
for
coming
back
to
and
to
speak
about
yahoo,
japan,
we
have
a
very
special
relationship
with
them.
They've
been
very
much
involved,
but
very
early.
The
Cyclops
could
have
a
creation
and
they
are
one
of
the
core
contributors.
Some
of
the
features
like
share
commands
they
completely
designed
and
implemented
it
and
yet
another
feature
which
comes
to
my
mind
is
branch
filtering,
so
yeah
puja
can
draw
that
feature
guided
by
design
and
supply
issues.
We
work
on.
We
create
issues
in
our
public
github
account
like.
C
C
Okay,
so
this
is
the
Kanban
board
which
I
was
talking
about
earlier,
where
we
collaborate
our
development
right.
So
we
have
a
backlog.
All
these
issues
come
for
more
open
scores,
so
we
design
everything
in
open
source,
so
nothing
is
designed
or
created
in
a
way
that
that
is
specific
for
any
one
specific
organization.
So
any
of
the
features
we
do,
we
make
sure
that
it's
done
in
a
way
that
it
thought
the
open
sourcing
right.
So
this
is
one
case
of
a
new
feature
recently
completely
driven
from
Yahoo
Japan.
C
So
this
was
a
feature
we
need
to
trigger
a
pipeline
based
on
the
github
release
tax.
So
our
typical
workflow,
like
works
like
this,
so
any
open
source
user.
They
come
and
create
an
issue
in
our
backlog
right
and
in
here
then
we
start
the
discussion
on
the
design
stuff
and
ultimately,
after
all,
the
inputs
are
taken.
We
prioritize
it
and
make
sure
that
it's
brought
up
in
this
board
the
top
of
the
backlog,
and
then
someone
will
start
pick
up
the
issue
and
start
implementing.
C
So
in
this
case,
all
these
are
called
all
these
contributions
are
made
from
the
Yahoo
Japan
team
and,
and
they
completely
walked
up
on
the
feature,
made
the
commits
and
pull
request
and
accepted,
and
we
have
seen
lot
of
few
protocols
which
came
in
from
external
users
as
well.
For
example.
Recently
there
was
an
issue
with
big
bucket.
We
have
integration
with
bucket
as
well.
There
was
one
user
who
was
using
bit
bucket,
but
this
was
not
used
within
our
organization
right.
C
B
Children,
let
me
you
know
I
think
you're
also.
You
happen
to
also
be
answering
the
second
question
which
I
have
to
do
around
governance
and
in
our
open
source,
friendly
model.
You
know,
I,
guess
one
of
the
questions
that
came
up.
Is
you
know?
How
do
we
govern
this?
How
do
we
operate?
Are
we
like
total
control?
B
Do
the
folks,
adora
collaborators,
let's
say
folks
in
Yahoo
Japan
collaborate
with
us:
do
they
have
admin
access
you
know,
could
they
could
they
add
to
the
project
as
well
and
are
we
looking
as
we
expand
this
into
the
foundation?
Are
we
receptive
to
having
other
parties
come
in
and
and
help
us
govern
this
project
so.
C
So
so
I
think
totally
Yahoo
Japan
has
admin
access
to
our
repositories,
so
they
can
create,
quote
requests
they
can
they
can
merge
it.
So
the
way
we
want
to
go
about
is
exactly
like
how
this
user
came
about.
He
had
a
problem
and
he
wanted
a
fix,
so
he
gave
the
code
contribution
now
how
we
want
to
go
about
this.
If
this
user
says
ok,
I
want
to
be
part
of,
this
could
have
a
journey
and
keep
contributing
code.
D
Just
want
also
this
video
I
just
also
add
that
currently
we
have
a
biweekly
meeting
that
goes
along
with
Japan
and
us,
and
that
meeting
can
totally
open
up
to
the
entire
open-source
community,
where
we
can
collectively
discuss
backlog.
We
are
at
ease
a
go
forward
so
that
making
that
meeting
public
can
be
done
and
as
soon
as
another
left
from
the
next
meeting
itself.
So.
B
Thank
you.
There
was
one.
There
was
one
last
question
on
the
chair
and
then
I'll
open
it
up.
If
there's
any
questions
that
have
come
up
subsequently,
but
the
last
question
on
the
channel
and
actually
I'll
take
it
I
think
it
really
had
to
do
with
how
we
as
an
industry,
can
ensure
that
you
know
we're
not
duplicating
efforts.
B
I
mean
we
spent
a
lot
of
time
at
Yahoo
and
now
cross
Verizon
media
building
this
out,
I
think
that's
precisely
why
we're
coming
to
this
foundation
I
mean
we're
so
glad
that
the
foundation
was
recently
formed
when
we
saw
we
immediately
reached
out
and
said
you
know,
we
also
don't
want
to
see
everyone
creating.
You
know
similar
similar
solutions.
You
know
for
the
same
problem.
B
What
we
want
to
do
is
take
the
solutions
that
we
built,
donate
them
to
the
foundation
and
work
with
our
peers
across
the
industry,
so
that
each
of
us
can
focus
on
solving
part
of
the
problem,
and
then
you
know
build
a
consolidated,
neutral
expanding.
You
know
a
platform
that
we
can
all
use,
so
we're
we're
actually
here
because
we
get
it
like.
B
We
want
to
make
sure
that
we're
not
building
something
that
locks
us
out
of
great
technologies
that
are
coming
out
of
our
pure
companies,
and
we
also
want
to
make
sure
there
are
the
technologies
that
we
built
and
that
we're
using
and
that
others
could
be
used
across
our
peer
companies
too,
because
we
leave
an
open
source
and
that's
I
mean
that's
really.
Why
we're
here.
So
if
any,
you
know,
if
anyone
else
I'm
from
the
foundation
or
listening
in,
has
questions
that
either
are
dropped
in
the
chat
or
want
to
pipe.
In
now.
G
Have
a
question:
this
is
Andy
yeah
all
right.
You
made
the
statement
that
this
was
you
know
an
open
source
project
and
that
you
built
nothing
specific
for
Verizon
media
mm-hmm.
So
it
goes
back
to
I.
Think
one
of
the
core
questions
of.
Why
does
Verizon
media
then
invest
in
this
project
if
they
get
nothing
specific
for
them?
So
it's.
C
It's
not
that
we
are
not
getting
nothing
specific
out
of
it,
so
this
project
is
currently
handling
daily
business
within
Verizon
media.
So
what
this?
What
we
have
proved
is
this
project
has
been
necessary
feature
set
and
the
scalability
that
it
can
handle
bills
at
that
scale,
and
we
haven't
built
this
with
any
technical
dependency
within
a
specific
product
within
the
licen
media,
but
that
doesn't
mean
that
it
doesn't
use
technologies
or
or
specific
thing
from
Verizon
media.
So
as
I
described
earlier,
SCADA
is
flexible
enough
that
it
can.
B
G
B
We'd
like
to
use
are
very
generic
built
tool
to
build
your
environments.
They
looked
at
and
it's
like
yeah.
This
is
completely
different
than
what
we
have,
but
it's
very
flexible
and
scalable.
Therefore,
you
know
the
folks
at
AOL,
who
are
not
part
of
our
company,
also
use
it
and
and
based
on
our
experience.
Expanding
to
the
AOL
builds
that
we
do
as
well
as
Yahoo
Japan
and
a
few
other.
A
Question
so
you
said
this
in
open
source
is
2016,
so
it's
been
three
years
like.
What's
the
experience
of
you
know,
I
guess
what
was
the?
How
was
that?
How
was
it
like?
What's
the
adduction
ever?
What's
the
I
guess
that,
presumably
a
big
possibility,
you
know
get
more
a
adduction,
but
also
more
contributions
like
did
that
he'd
mentioned
like
a
Yahoo
Japan
is
a
big
contributor.
A
B
B
It
was
really,
let's
make
sure
that
it's
open
source,
let's
make
sure
that
we
haven't
locked
anything
in
that
we
could
work
with
other
companies
and
that,
as
we
had
opportunities
that
we
were
able
to
bring
up
to
really
leverage
those
opportunities
without
locking
we
were
also
going
through
as
a
company
on
a
transition
in
2016
or
2017,
as
as
the
company
itself
was
sold.
So
you
know
the
objective,
really
wasn't:
let's
get
as
many
external
contributors
as
we
can.
B
The
objective
was:
let's
make
sure
that
when
the
time
is
right,
we'll
be
able
to
do
that,
we
think
the
time
is
right.
Now
we
think
the
time
now
that
we
have
the
CD
foundation
as
as
really
a
shared
resource
in
the
industry.
We
think
we're
ready
now
to
expand
the
the
community
that
we
have
and
really
build
on
what
we
have
demonstrated
in
the
past.
To
say
great,
you
know
we
have
a
great
relationship
with
a
few
partners.
We
would
like
a
great
relationship
with
even
more
partners.
C
And
from
what
will
open
source
CV
flak
channel,
we
have
seen
folks
from
a
couple
of
different
organizations
coming
and
asking
postings
on
okay.
How
do
how
do
I
do
this,
and
that's
also
potential
I,
saw
that
I
use
else
out
there
who
are
running
forever
within
their
organization,
but
right
now
we
are
not
actively
clacking
it,
but
we
want
to
do
more
on
building
those
collaborations.
A
B
E
D
The
team-
and
we
are
kind
of
going
all
out
in
terms
of
how
the
the
products
that
we
have
used
internally
at
scale,
can,
we
can
be
made
sure,
can
be
made
available
for
all
the
users
outside.
At
this
point,
this
quality
is
available
for
users
to
contribute
use
it.
There
is
no.
This
is
pretty
much
it
is.
It
is
as
good
as
any
any
open
source
library
we.
We
are
not
in
the
business
of
any
enterprise
solutions
or
selling
or
accepts
just.
E
D
Saw
this
software
work
for
us
at
scale,
high
performance
extremely
secure,
and
we
believe
that
the
community
can
can
really
benefit
from
it,
and
we
also
saw
the
leadership
at
Citi.
Foundation
was
a
great
opportunity,
so
we
believe
that
Citi
foundation
can
also
enable
and
provide
platforms
and
opportunities
for
this
software
to
be
benefited
for
all
the
cambric.
There's.
E
E
B
Okay,
so,
as
you
see,
we
have,
we've
got
a
site,
it's
full,
you
know
the
code
is,
is
documented.
We
have
a
guide
and
what
we're
really
looking
for
is
its
partnership
with
the
foundation.
I
mean
we're
again
we're
excited
that
there
was
this
this
foundation.
We
think
that
there's
this
great
opportunity
for
us
to
work
with
the
foundation.
A
Yeah
I
think
it's
I
mean
I,
think
it's
great
that
them
I
mean
one
of
T.
My
what
I
was
really
hoping
through.
The
CDF
is
like
we
can
be.
Are
a
group
of
people
who
are
passionate
about
solving
the
problem
in
this
space,
can
check
the
join
hands
together
and
then
make
a
bigger
impact
collectively
and
that
sort
of
like
I,
think
that
really
propelled
open-source
and
so
I'm,
not
loving,
loving
the
hobby.
A
You
guys,
you
know
they
show
up
and
they're
here
there
is
seems
really
great
and
also
the
convo
I
think
what
the
the
questions
that
this
can
an
A+
underscores
for
me,
which
you
know,
has
being
kind
of
like
a
problem
all
along
well,
the
challenge
all
along
that
needs
to
be
solved
was
high,
the
high
to
actually
join
during
times
more
effectively
at
the
pasta.
You
know
like
we
all
bring
our
own
toys
and
I
got
trying
a
different.
A
You
know
play
play
play
in
the
different
you
know
like
so
well,
when
I
have
my
kids
play
with
other
friends
like
sometimes
they
are
the
same
around
the
same
people
and
they're
all
playing
with
their
cell
phones.
They
are
not
talking
they're
interacting,
and
that
would
be
a
shame,
I
think
so
you
know
that's
something.
A
It
wasn't
too
much
trouble
to
kind
of
make
them
and
work
on
the
same
thing
by
abandoning
some
of
the
code,
but
here
the
amount
of
things
that
we
collectively
have
done
in
our
own
cell
phone.
So
to
speak,
it's
pretty
big
I.
Actually
like
there's
different
approaches
needed,
but
I,
don't
you
know
I,
don't
so
that's
something
I
feel
like
we.
H
A
I
B
Well,
we
first
wanted
to
share
that,
so
we
have
the
like
a
questionnaire.
We
have
I
think
a
lot
of
it
answer
and
we
really
wanted
to
first
present
some
more
details
so
that
the
foundation
knew
what
it
is
that
we
were
talking
about
in
the
scope.
You
know
how
well-built
and
thought-out
it
is
so
that,
when
you
receive
the
proposal,
you'll
have
a
little
more
context
about
it,
but
we'll
take
that
next
step,
we'll
submit
the
proposal
and
just
know
we're
here
to
be
we're
here
to
partner
and
I.
B
A
I
C
B
Yeah,
that's
that
listen.
The
team
was
great.
So
if
there
are
any
other
questions
from
the
folks,
I
mean
we've.
We've
taken
we've
taken
a
bit
of
time,
but
there's
everything's
open
writings
on
github
everything's
documented.
So
please,
please
do
take
a
look
at
it
reach
out
to
us,
and
you
know
let
us
know
how
we're
doing
with
respect
to
our
proposal
and
our
next
steps
so
that
we
can
continue
to
work
together.
But
again,
thank
you
for
for
upright
or
taking
the
time
to
to
hear
us
out.
H
G
H
So
I,
just
last
week
we
discussed
the
CDF
working
group
proposal
that
I
started
and
the
main
feedback
was
that
it
would
be
easier
to
review
if
we
included
the
full
scope
of
also
the
CDF
C's
I
might
intentionally
left
that
out
the
first
time,
but
they've
done
a
feedback.
I
added
another
proposal
around
it.
So
it's
easier
to
compare
when
one
it
would
be
appropriate
versus
the
other
for
different
efforts
going
on.
A
H
So
I'll
give
people
a
couple
more
days.
Then
we
have
a
holiday
coming
up
to
add
comments
here
to
the
stock.
After
that,
I
will
probably
early
next
week,
turn
us
into
a
pull
request
and
markdown,
and
then
we
wrap
it
up.
Github
yeah
right
here.
So
if
you
haven't
gotten
a
chance
to
take
a
look
up,
please
do
soon.
H
I
just
did
right,
yeah
all
right
and
then
I
think
I'm,
not
sure,
if
she's
in
the
call,
but
she
actually
went
through
and
put
together
a
prototype
proposal
for
a
security
working
great
securities
because
she
wants
to
start.
So.
If
we
had
those
side-by-side,
we
can
kind
of
do
a
beta
test
of
how
this
works
out
at
Agnes.
I
am.