►
Description
For more Continuous Delivery Foundation content, check out our blog: https://cd.foundation/blog/
B
A
I
did
get
an
email
about
accessing
the
speaker
portal,
but
I
think
it.
I
was
a
bit
confused
because
I
didn't
get
the
notification
about
being
accepted
or
not.
C
Yeah,
I
think
that
we
got
the
talk
accepted.
Maybe
it
was
just
the
email
wasn't
only
to
me.
Let
me
see
if
I
can
forward
that
to
you.
A
E
C
Andrea,
can
you
share
your
personal
email
with
with
me
in
the
slack,
for
some
reason,
I
don't
have
it
in
my
gmail.
I
should,
but
it's
not
there.
C
Thank
you,
everyone
for
joining-
let's
definitely
get
started
with
this,
so
we
can
keep
it
like
short
and
simple.
So
I've
shared
a
link
for
the
vocabulary
agenda
today.
If
someone
has
any
topic
for
the
agenda,
please
do
add
there
to
to
that
document.
It's
in
the
zoom
chat.
I
will
just
paste
it
again
and
let
me
see
if
I
can
share
my
screen
quickly,
so
we
all
see
the
same
document
here.
We
are
so
I've
added
a
couple
of
topics,
and
I
think
that
we
can
get
rid
of
this
one.
C
Just
a
quick
update
for
everyone.
We
have
like
merged
the
the
gold
library.
So
this
is
the
goal
library
that
it's
basically
inside
the
c
events
cde
sdk
go
so
that's
kind
of
like
where
the
library
like
the
cloud
events,
library,
leaves
for
now
it's
early
days,
so
it's
pretty
hacky
and
we
expect
to
be
able
to
improve
it
like
adding
tests
and
adding
like
removing
code
duplication
and
making
it
a
little
bit
better
like
by
doing
the
poc
that
we
are
planning.
C
I
don't
know
if
everyone
in
the
call
is
aware,
but
we
are
building
a
plc
for
the
devops
world
and
also
for
silicon,
which
is
next
week.
I
think
so.
We
need
to
definitely
get
cracking
on,
and
today
I
had
a
meeting
with
jurgen
and
joe
harness
from
captain
to
get
started
on
building.
Basically,
what
is
called
a
captain
translator,
I
think
that's
basically,
johannes,
has
a
diagram
there.
C
That
is
pretty
useful,
so
I
think
that
we
will
start
using
those
names
for
now,
johannes,
if
you
can
add
a
link
to
that
diagram
or
if
you
can,
you
know,
share
it
and
add
it
into
a
pull
request
into
the
repository.
I
think
that
will
be
pretty
interesting.
D
Yes,
of
course
I
can
do.
I
was
just
waiting
for
getting
the
first
pr
merged,
but
now
I
can
submit
a
change
and,
as
we
just
discussed
already
in
the
chat
yeah,
I
moved
the
tekton
controller
out
of
this
yellow
rectangle,
because
it
should
not
be
part
of
that
component.
That
takes
care
of
translating
cdf
events
into
cabinet
events
sounds.
C
Good
sounds
good,
sounds
good,
so,
basically
the
plc
for
everyone
to
be
aware.
I
think
that
it's
going
to
be
composed
by
these
components.
At
least
these
first
two
I
would
say
the
last
one
is
definitely
optional
and
I
don't.
I
will
not
focus
on
that
for
now.
I
guess
that
until
next
week
that's
not
going
to
happen,
but
I
know
that
there
is
some
people
working
also
on
the
jenkins
side,
building
a
plugin
for
cloud
events,
so
in
you
know
in
later
iterations.
We
can
definitely
include
that.
C
I
wanted
to
check
with
andrea
and
I
don't
know
if
viva
is
not
here.
I
don't
know
if
he's
here,
no
he's
not
yep
go
ahead.
A
No,
I
think
it's
a
bit
too
late
for
him.
You
have
to
join,
but
I've
been
in
touch
with
him.
A
A
A
C
I
think
that
it's
it's
a
good
time
to
today
kind
of
like
what
we
were
think
thinking.
Thinking
about,
like
the
captain,
translator,
like
the
captain
side
and
basically,
what
we
discussed
is
kind
of
like
a
refinement
of
what
we
have
been
discussing
in
like
previously,
without
having
too
many
details
now,.
C
This
is
what
I
would
be
expecting
from
the
tech
consider,
and
this
is
something
that
I
want
to
check
with
andrea.
Is
that
to
be
able
also
to
be
able
to
consume
cd
events.
Is
that
something
that
can
be
hacked
at
least
for
next
week?.
A
Yeah,
I
think
so
that
goes
through
a
component
in
tecton
called
triggers
and
that
receives
generic
post
over
http
and
then
it
you
can.
You
can
use
basic,
basically
json
path,
to
extract
information
from
the
json
body.
So
that's
yeah,
that's
what
we
do
for
tekton
events
as
well,
so
we
just
need
to
create
like
the
the
binding
for
for
this
kind
of
events
and
then
but
it's
it's
ammo
based.
So
there's
no
go.
A
Yeah,
basically
detecting
triggers
you,
you
create
a
crd,
called
event
listener
that
sets
sets
up
a
sync,
so
it
gives
you
a
url,
and
then
you
post
the
the
cloud
event
to
that
url
and
then
contact
one
side.
You
set
up
some
rules,
filtering
rules
and
to
say:
okay,
yeah
accept
this
event
and
take
ex
extract
this
information
for
this
event
and
start
fighting
with
that.
F
A
F
A
A
It
depends
how
you
want
to
expose
it.
So
if
you
put
in
an
ingress
on
top
it's
a
variable
publicly
and
yeah,
so
by
default,
there's
no
there's
no
authentication
mechanism
in
build,
but
like
for
for
github,
for
example,
github
events-
you
can
have
a
secret,
so
we
have
a
specific
component
to
validate
the
secret
that
comes
from
github
events.
F
Yeah,
I
was
thinking
because
we
have
to
somehow
send
also
the
cloud
event
from
captain
via
the
translator
as
a
as
a
cd
event
than
to
detect
them.
I
guess,
and
it
has
there
has
to
be
somewhere
an
endpoint
where
we
can
send
it
to.
C
F
F
C
C
So
yeah
yeah.
A
C
In
one,
my
experience
so
far
with
with
them
with
captain
is
that
they
are,
you
guys,
are
installing
a
bunch
of
components
and
they
require
some
specific
configurations
at
least
looking
at
the
quick
start.
I
didn't
like
when
I
looked
at
the
quest,
quick
start
and
and
saw
the
installed
components.
I
basically
said
yeah.
I
can
probably
get
this
working
on
kind,
but
I
will
not
spend
time
now
because
you
know,
because
the
quick
start
basically
set
up
all
the
components
for
me
and
all
the
routes
and
also
exposes
the
things
outside.
C
I
can
give
it
a
try
to
to
run
tecton
in
there
and
see
if
it
kind
of
works
out
of
the
box,
I'm
guessing
that
it
will.
But
yeah
do.
We
need
k
native
for
something
right
now,
if
we
don't
use
the
brokers,
for
example,
or
do
we
need
it
for
the.
A
C
Good
stuff,
andrea,
so
basically,
what
I
think
that
we
should
build
is
that
if
we
build
a
pipeline
that
basically
you
know
creates
a
contain-
I
mean
I
don't
know
if
you
have
kind
of,
like
example
pipelines,
but
basically,
if
we
create
a
pipeline
that
says
it
doesn't
need
to,
it
needs
to
only
emit
an
event
saying
there's
a
new
artifact.
C
C
Put
it
running
just
yet,
but
I'm
pretty
close,
and
if
we
do
that,
then
basically,
what
captain
is
going
to
send
us
back
is
a
captain
event
saying
you
know
that
you
know
a
pipeline
is
to
be
triggered
in
tecton,
and
I
will
translate
that
back
to
a
cd
event
that
you
should
be
able
to
pick
up
with
that
trigger
in
that
trigger.
So
I
would
just
send
you
that
into
a
trigger
and
we
should
be
able
to
trigger
a
second
pipeline
from
there.
C
C
C
D
I
think
the
use
case-wise
it
makes
totally
sense
I've
just
one
technical
requirement
that
may
come
up
when
when
we
go
a
little
bit
down
the
road
because
to
close
or
to
to
tell
captain
that,
for
example,
a
deployment
event
is
finished,
captain
needs
a
reference
which
trigger
was
used
and
this
information,
it's
just
a
triggered
id
needs
to
be
stored
somewhere
and
now.
D
My
question
is
whether
we
can
pass
it
to
tacton
and
tekton,
then
als
so
returns
it
back
to
the
translation
layer
so
that
the
translation
layer
knows
which
event
to
reference
and
finally
forwards
it
to
captain,
so
that
then
captain
can
make
a
decision.
Okay,
the
deployment
is
done.
This
is
one
technical
issue
that
might
come
up,
but
apart
from
that,
the
use
case
makes
sense,
and
it's
totally.
A
Feasible
think
from
tecton
side
shouldn't
be
a
problem,
I'm
just
wondering
where,
where
we
store
this
into
the
city
into
the
cde
protocol,.
C
A
So
what
I,
what
I
can
do
in
the
what
I
can
do
on
on
tekken
side,
I
mean
I
can
build
the
cd
events
with
the
headers
set
as
needed,
and
then
in
the
body
I
could
put
like
the
entire
pipeline
techno
pipeline,
which
will
include
somewhere
this
id,
but
then
you
need
to
know
where
to
go
and
extract
it.
The
alternative
would
be
to
to
define
some
other
body
or
header
where
we
we
set
this
whatever
is
easier.
I
guess.
A
Yeah,
do
you
have
an
example?
Jonas,
maybe.
D
D
Let
me
just
yeah
here
we
have
it.
D
All
right
here
I
have
on
the
right
side.
You
can
see
a
sequence
execution
as
orchestrated
by
captain
ian,
for
example,
a
deaf
environment
and
everything
gets
kicked
on
by
a
sequence
triggered
event.
The
this
is
the
the
initial
initial
event
that
gets
sent
to
captain
after
a
build
happened
and
another
artifact
is
available
and.
F
Maybe
I
I
just
have
to
clarify
here,
I
think
in
the
last
meeting.
We
also,
we
always
called
it
a
new
artifact
event,
but
it's
actually
not
called
the
new
artifact
event.
It's
more
a
triggering
of
some
deliberate
sequence.
So
in
the
pre
in
the
first
version
of
captain
it
was
called
the
new
artifact,
but
it's
now
more
generic
and
we
just
call
it
the
triggering
of
a
sequence
that
can
be
triggered
because
of
a
new
artifact.
But
this
is
where
we
we
jump
into
the
captain
execution.
D
Then
captain
yeah
takes
a
look
into
a
process
definition
which
step
to
execute,
and
the
first
one
in
this
case
is
a
deployment
event
and
captain
bill
sent
out
a
so-called
sh
captain
event
deployment
triggered
just
to
inform
then
the
service
that
is
listening
to
this
event,
that
this
task
is
now
going
to
be
started,
or
you
should
do
your
job
now
and
this
event
also
has
a
unique
id
it's
this
one.
D
This
is
yeah
unique
and
when
then
some
extension
executes
upon
this
event,
then
they
send
back
a
so-called
finished
event,
sh
captain
event
deployment
finished
and
they
tell
us
which
triggered
id.
They
are
referencing
and
this
id
we
need
to
close
or
to
make
to
make
this
step
to
close
the
step
and
to
make
the
decision
that
everything
is
now
finished
and
we
go
on
with
the
next
task
in
the
sequence
and
and
do
yeah
continue,
and
this
id
I'm
a
little
bit
worried
about
this
id.
D
D
D
C
In
this
case,
what
I'm
thinking
is
that
you
know
tecton
is
going
to
be
running
some
pipelines,
sending
some
cd
events
without
this
right.
The
first
kind,
like
the
first
event,
that
we're
going
to
send
it's
not
going
to
have
this
one
right.
This
is
just
the
the
second
time
or
the
time
that,
when,
when
we
get
that
from
from
captain.
D
Right
true,
the
first
one
doesn't
does
not
have
this
event.
This
id.
This
isn't
totally
correct.
It
just
has
a
unique
id,
but
this
can
be
any
arbitrary
number,
but
also.
D
A
That's
very
interesting,
I
think
yeah
we
might
need
to
think
about
some
general,
maybe
context
field
in
the
protocol
that
could
be
used
by
different
application
for
transporting
this
kind
of
information.
E
I
think
what
we're
talking
about
now
is
actually
how
how
to
link
events
together,
how
to
relate
them
to
each
other
right,
either
through
a
common
context
or
through
through
links
between
the
different
events
that
we
say
are
related
in
some
way.
E
So
I
think
that
the
way
to
do
this
in
a
generic
way
andrea
would
be
to
either
have
a
context
or
and
or
I
would
say,
event
references
between
the
events
like
links
that
we
call
them
in
eiffel.
E
A
A
C
I
think
that
we
can
just
handle
that
logic
in
the
translation
phase
and
then
see
if
we
can
provide
it
to
tecton
and
then
reuse
it
whenever
we
get
events
from
tecton
and
just
extract
it
and
generate
the
new
events
for
for
captain.
In
this
case,.
F
Is
there
a
generic
field
in
in
the
events
that
where
we
can
just
put
in
like
new
information,
let's
say
here
in
the
in
the
captain,
events
in
the
data
block
anything
can
go
into
the
data,
so
you
can
just
add
anything
else.
Let's
say
here
we
have
a
deployment
or
project,
but
if
it
should
be
first
name
and
last
name,
you
can
just
edit
there.
It's
not
it's
not
specified
and
if
there
is
also
a
way
how
to
just
define
the
payload,
then
maybe
like
for
a
temporary
solution.
F
We
put
in
also
the
context
data
into
the
payload,
and
maybe
then
we
can
kind
of
loop
it
through
or
kind
of
yeah
kind
of
carry
it
with
all
the
events,
and
then
we
just
have
to
make
sure
that
we
don't
lose
the
captain
context
the
triggered
id
these
kind
of
things
we
have
to
take
care
of,
but
right
now
it's
a
technical
implementation.
B
C
Not
right
now
I
will
be
definitely
you
know
writing
like
trying
to
ask
questions
as
they
come.
Basically,
I
will
try
to
work
with
the
authentication
thing.
I
think
that's
almost
done
yeah
receiving
cd
events
that
probably
are
going
to
come
from
tecton
and
I'm
guessing
that
from
the
tecton
controller
I
can
I
you
can
define
kind
of
like
a
sync
for
that.
I
have
a
service,
basically,
where
you're
going
to
be
sending
the
events,
I'm
guessing
that
that's
kind
of
like
how
it
will
work,
and
I
will
definitely
ping
your
johann.
C
And
jurgen,
if
I
have
any
questions
because
yeah,
I
think
that
I
will
just
be
working
tomorrow
on
on
at
least
making
sure
that
captain
can
receive
the
events
and
then
that
I'm
formatting
in
the
right
way.
So
we
can
actually
see
something
in
the
in
the
user
interface,
as
I
mentioned
before,
I'm
not
entirely
sure.
What
kind
of
configurations
do
I
need
to
do
in
captain
itself
to,
for
example,
configure
the
sequences.
D
Okay,
let
me
just
show
you.
D
D
They
have
a
dev
stage
staging
and
a
production
stage,
and
then
within
each
stage
you
have
to
you
can
configure
our
sequences
and
when
we
take
a
look
at
the
delivery
sequence,
this
one
is
then
composed
of
a
couple
of
tasks,
and
this
is
then
the
process
that
or
the
the
sequence
of
tasks
that
gets
executed
and
orchestrated
back
up,
meaning
that
deployment
happens.
First
then
test
evaluation,
release
and
so
on,
and
this
is
the
way
that
you
configure
your
captain
project
and
then
to
set
up
the
sequence.
You
want
to
execute.
C
Yes,
I
I
and
then
what
do
I
do
with
that
file.
So
basically
I
just
apply
that
and
it
will
pop
up
in
in
the
captain
user
interface.
D
True
yeah,
you
apply
that
file
or
you
creating
a
captain
project.
You
just
reference
a
shipyard.yaml
or
any.
A
C
Well,
yeah,
I
thank
you
for
some
clarifications
there.
Whenever
I
try
to
do
that,
I
will
let
you
know,
and
probably
I
will
be
needing
some
help
just
to
make
sure
that
I
just
start
in
the
in
the
you
know
with
the
right
step
and
not
not
not
going
into
like
a
completely
wrong
direction
and.
D
C
Time
on
things
that
you
can
probably
solve
quite
quickly:
okay,.
C
A
That
inside,
if
you
want,
I
mean
we
just
need
to,
I
guess-
have
a
sample
app
in
a
docker
file.
If
anyone
has
any
one
that
would
like
to
use
that
we
can
build
and
deploy
easily
or
I
don't
know,
I
can
pick
one
and
then
basically,
I
can
set
up
two
pipelines,
one
for
the
build,
which
is
pretty
straightforward
and
the
second
one
for
the
deploy,
and
I
think
the
main
thing
for
the
deploy
one
is
to
make
sure
we
have
parameter
to
capture,
to
capture
the
context
that
captain
needs
right.
A
And
if
we
output
that
context
back
as
pipeline
results,
then
you
should
be
able
to
translate
those
back
into
yep
into
the
cd
format.
A
A
Okay,
it
would
be
nicer
if
it's
in
the
cde
format,
but
if
you're
short
on
time,
we
could,
for
the
first
plc,
say,
send
it
in
the
yeah
in
the
captain
format,
because
from
from
from
techno
point
of
view,
as
long
as
it's
http
with
json
attached,
it
doesn't
matter
you.
I
just
need
to
write
some
json
path
rules
and
then
you
can
extract
the
information
in
it
from
where
I
need
it.
C
A
Okay,
the
other
thing
we
have
a
poc
folder
now
on
the
repo
where
the
the
initial
version
of
the
diagram
is
so
I
think
we
could
use
some
folders
in
there
to
store
the
different
resources,
so
I
would
create
like
the
techno
pipeline
yaml
in
that
folder
in
the
tectum
folder
and
so
forth.
We
could
have
like
a
folder
for
the
captain
resources.
So
we
have
all
the
ammos
together
in
one
place,
sounds
really
good.
Yep.
C
Sounds
really
good,
let's
do
that.
So
if
I
create
any
jumble
files,
I
will
just
push
them
there
and
I
have
another
repository
for
the
translator
component
and
I
will
have
another
repository
for
the
service
like
the
captain
service,
which
is
basically
the
one
picking
up
events
from
from
captain
and
forwarding
stuff
to
tecton.
So
there
are
like
two
separate
components
that
we
can
merge
in
the
future.
That's
something
that
we
discussed
late
before,
but
yeah.
C
A
Out:
okay
and
that's
that's
a
configuration
on
captain
side
or
is
a
chill
service
company
service.
C
C
So,
let's
keep
it
simple
for
now.
You
know,
I
know
that
it
sounds
complex,
but
I
will
just
create
two
very
very
simple.
You
know
goal
projects.
I
will
make
the
connections.
I
will
test
them
to
see
the
work
and
as
soon
as
I
have
that
I
will
definitely
start
pinging
you
andrea,
to
see
if
I
can
install
techno
and
just
get
everything
in
one
place.
A
Yes
great
as
soon
as
you
have
one
example
of
the
cloud
events
that
detector
would
be
receiving.
C
Yeah,
I
know
yep,
that's
that's
one
that
I'm
fully
aware,
but
yeah.
I
have
so
many
stuff
to
do
this
week.
So
I
just
want
to
make
sure
that
I
plan
to
have
enough
time
to
finish
kinda
like
the
first
version
of
this,
but
I'm
guessing
that
because
of
the
urgency
of
this
tomorrow,
I
will
be
working
on
this
full
day
and
thursday.
I
might
be
able
to
do
the
same,
so
I
yeah,
I
will
just
keep
bringing
people
and
you
know
and
asking
for
help
all
right,
good
stuff.
C
We
do
not
have
any
slides
for
any
of
this
stuff,
so
the
more
that
we
can
add
to
the
plc
with
diagrams
and
explanations.
Now
that
we
can
like
discuss
the
use
case,
I
think
that
we
are
all
kind
of
like
aware
of
how
it
should
work
if
people
can
help
like
feeling
feeling
kind
of.
Like
the
you
know,
the
repository
like
the
readme
and
all
that
stuff
explaining
what
we
are,
what
we
are
trying
to
achieve.
That
will
be
super
helpful
to
create
then
a
presentation,
for
example,.
A
B
B
C
I
don't
know,
I
think
that
we
need
to
basically
record
the
presentation
for
devops
world
and
I
will
check
the
deadlines.
Probably
let
me
do
that
tomorrow,
just
to
make
sure
that
we
have
enough
time,
usually
if
they
do
recordings,
they
have
like
a
week
to
do
the
recordings,
and
I'm
assuming
now
that
that's
the
the
you
know
the
more
the
more
strict
deadline
that
we
will
have,
because
we
will
need
to
have
a
presentation
and
being
able
to
record
that
with
the
demo
yeah.
B
So
I'm
almost
afraid
to
open
my
mouth,
because
I
have
a
tendency
in
this
group
to
not
deliver
what
I
commit
to,
but
I
do
work
a
lot
with
presentations,
so
I
will
pinch
myself
really
hard
and
really
try
to
help
with
that
part
and
and
I'll
set
up
some
expectations
on
myself
and
maybe
deliver
this
time.
That
would
be
nice.
C
Thank
you,
eric
that's
highly
appreciated
and
again,
maybe
like
the
presentation
itself,
is
one
thing
that
we
can
start
working
together
based
on
the
content
on
the
on
the
repository
right
like
so.
If
you
have
like
a
plc,
read
me
there.
Let's
add
more
content
in
there
and
then
creating
like
the
slides.
It's
pretty
straightforward.
B
C
F
B
Great
you
can
make
some
pr
for
the
birth
of
better
session,
because
there
are
apparently
four
of
them
going
on
in
parallel,
and
we,
of
course
want
people
to
join
the
right
one.
Certainly,
yes,.
C
C
Okay,
fellas,
I
will
definitely
jump
into
coding
mode
now.
So
if
there
are
no
other
topics
to
discuss,
you
know
we
can
always
keep
chatting
into
slack
yep.