►
From YouTube: CDF - SIG Events Meeting - 2021.09.13
Description
For more Continuous Delivery Foundation content, check out our blog: https://cd.foundation/blog/
A
C
A
C
C
C
Okay,
so
I
guess
we
should
not
wait
anymore
yeah,
so
welcome
here
today,
as
I
think
you
might
have
seen
in
the
mail
I
sent
out
last
week
today
we
are,
we
are
visited
by
derek
wang
from
argo
arlenz
actually,
and
he
will
present
what
arguments
is
just
followed
by
maybe
a
short
discussion
on
how
we
can
continue
our
cooperation
and
then
we
will
continue
on
discussing
on
our
the
name
of
our
new
event
protocol,
let's
see
where
we
can
get
today.
C
C
Yeah,
the
only
one
was
that
I
should
prepare
for
the
name
setting
to
get
with
andrea
and
we've
done
some
preparation
there.
So
we'll
come
to
it
later
on
yeah,
anything
from
you,
andrea
or
should
we
go
ahead?
C
I
think
for
me
yeah,
okay,
so
let's
go
on
then
and
derek
the
stage
is
yours.
Thank
you.
Let
me
share
my
screen.
D
All
right,
my
name,
is
derek
dewon,
I'm
from
argo
community.
My
major
worked
on
argo
events
and
oracle
workflows,
and
I'm
very
I'm
super
happy
to
be
in
one
here
to
give
introduction
about
argo
events,
and
I
hope
argo
events
could
be
an
extra
option
for
cd
community
for
events
handling
stuff.
D
Before
I
started,
I
just
want
to
give
a
brief
brief
introduction
about
the
organ
family.
We
actually
have
four
cnc
projects.
D
Stuff-
and
we
also
have
workflows,
which
used
for
workflow
manually,
used
by
ai
machine
learning,
guys
is
very
popular
there
and
we
have
events
we
eventually
used
to
handle
the
event
event.
The
events
related
to
stuff
argo
events.
What
is
arguably
events,
arguments
is
the
event-driven
workflow
automation
framework
and
it's
for
kubernetes,
and
you
can
use
arc
events
trigger
actions
based
on
the
events
you
received,
and
you
can
do
it
by
using
kubernetes
as
well.
D
Right
now,
arguments
supports
25,
different
type
of
events,
which
includes
those
queue-based
or
pub
sub-based
event:
sources
like
aws,
sqs,
kafka,
nas
or
wrapping
queue,
this
kind
of
stuff
and
also
encodes
like
there's
a
web-based
web-based
even
source
like
github,
gila,
bitbucket
or
just
a
regular
webhook
event
source.
D
It
also
can
support
like
those
file
based
event
source
like
f3
bucket,
which
is
a
menu
here
and
also
like
hdfs
or
just
local
file
based
events
and
arguments
also
can
do
like
watch
the
events
from
kubernetes
itself,
not
not
only
the
kubernetes
equivalent
events,
but
any
kubernetes
object
change.
You
can
use
argo
events
to
watch.
B
Derek
are
these
incoming
events
that
argo's
listening
for,
or
are
these
outgoing
events
that
argo
is
publishing?
Is
it
it's
incoming
message?
Okay,.
D
It
has
a
lambda
or
like
openwhisk,
which
is
a
function
based
action
can
take
and,
of
course
we
support
our
work
like
this.
You
know
same
family,
so
it's
first
class
citizen
and
you
also
can
do
like
send
the
message
to
like
cubase
or
pubsub-based
services
like
kafka,
nas
or
you
can
just
make
any
kind
of
http
request
which
enables
to
do
any
action
to
trigger
a
service
workload.
D
And
you
also
can
use
it
to
create
any
kind
of
kubernetes
resource,
no
matter
if
the
deployment,
the
configman
or
any
kind
of
crd
object,
if
you
could
not
find
any
event
source
or
actions
supporting
arc
events,
you,
actually
we
actually
provide
a
interface
for
you
to
implement
your
own
demand,
source
or
sensor
or
action.
You
want
to
take
as
we,
we
name
them
as
general
events
or
generate
trigger
kind
of
stuff.
We
we
provide
a
grpc
interface.
You
only
need
to
implement
the
jvc.
D
You
need
to
write
your
own
code
to
implement
jrpc
interface,
so
you
can
get
the
hook
to
arbitrary
system.
D
And
then,
let's
look
look
at
some.
The
key
concepts
in
oracle
events
were
the
first
crd
object.
I'm
going
to
introduce
is
even
source,
even
source
is
used
to
define
something
you
are
interested
in
typically
is
events
from
either
outside
of
the
cluster
or
inside
of
the
cluster
and
the
event
source
job
is
to
watch
those
events
and
then
send
those
events
to
an
internal,
even
bus.
D
Let's
look
at
some
examples
about
events
or
two
to
to
see
how
these
zero
objects
look.
Look
like
so
on.
The
left
side
is
a
event.
Source
object
named
a
webhook,
it
configures
events
named
example.
Here,
I'm
going
to
explain
later
and
this
human
source
this
event
listens
to
port
1200.
It
has
endpoint
example
and
accepts
a
message
like
post.
So
this
example,
the
example
on
the
left
is
a
single
event.
Computer
one
email
source
that
actually
even
source
supports
to
configure
multiple
events
in
one
even
source
object.
D
D
Configured
two
of
them
are
named
example,
one
example:
example,
two
there
are
calendar
type,
even
source
and
the
third
one
is
a
menial
type
which
can
be
used
for
like
three
events.
We
know
that
example.
Three,
I
didn't
give
the
detailed
configuration
here
just
give
you
some
concept
about
how
arguments
even
source
look
like,
and
then
we
look
into
the
second
most
important
concept.
In
our
events,
were
we
name
it
as
sensor?
D
A
sensor
is
used
to
listen
to
a
human
bus
for
certain
type
of
events,
and
then
you
trigger
your
actions
conditionally,
based
on
your
configuration.
A
sensor
typically
contains
two
parts.
The
first
part
declares
what
kind
of
dependent
what
kind
of
even
source
you're
interested
in
we
named
as
dependency
and
the
second
part
which
defines
what
kind
of
action
you
want
to
take,
and
we
name
this
like
trigger
so
trigger
is
an
action
you
want
to
take.
D
As
the
sensor
example,
so
in
sensor
first
part,
I
have
a
dependencies,
you
give
the
event
source
name
and
even
then,
and
give
a
dependency
name,
and
this
triggers
part
this
this
trigger
is
like
a
simple
http
trigger.
I
I,
when
I
see
an
event
from
sqs,
I
want
to
make
a
hd
request,
get
the
request
to
medium.com
it's
a
simple
example
of
sensor.
D
This
architecture,
diagram
of
arc
events.
So
when
I
use
arc
events
typically,
there
are
two
kinds
of
name:
space
are
involved.
The
first
is
system
namespace,
which
is
to
install
those
controllers.
We
have
even
source
controller
sensor,
controller
and
even
bus
controller.
Installs
are
the
names,
argo
events
namespace
the
other
kind
of
event.
Namespace
is
a
username
space.
D
D
So
when
event,
source
object
or
sensor
objects
created
the
event,
source,
controller
and
sensor
will
create
and
corresponding
parts
to
for
event
source.
It
creates
power
to
watch
the
events.
D
For
example,
you
have
sqs
the
there
will
be
a
event
source
power
watch,
sqs
event
and
also
the
sensor
object
will
have
a
sensor
pod,
which
is
which
is
there
to
be
ready
to
like
trigger
actions
but
the
event
source
when
the
event
source
part
see.
There's
events
coming,
it
does
not
send
the
events
directly
to
the
sensor.
Instead,
it's
sent
to
object,
name
even
bus,
so
here
we
have
even
even
bug
the
third
crd
object,
name
even
bus
here
in
the
system.
D
D
So
this
event
buzz
brings
two
benefits.
One
of
them
is
like
so
first
one
is
a
pop
star
system,
so
it
naturally
provided
pops
up
feature.
So
events
events
can
be
used
by
multiple
actions.
D
The
second
benefit
is
the
second
benefit
is
provides
a
kind
of
persistence,
persistent
message
used
by
other
events
used
by
the
system.
So
if
you
have
a,
for
example,
sqs
type
image
source,
it's
a
queue:
it's
not
a
pops
up.
D
Basically,
so
when
you
consume
the
message
from
sqsq,
you
have
to
normally
have
to
read
it
and
then
delete
the
message
from
the
queue
and
this
way
you
cannot
be
reused
by
any
other
kind
of
consumers,
but
by
using
argo
events
when
the
sqs
event
stores
written
messages,
execute
it
actually
send
to
an
internal
pop
sub
system,
and
then
it
could
be
reused
by
any
kinds
of
consumers
defining
the
sensor,
which
is
actions
right
now.
The
events
we,
if
you
create
a
eventbuzz
object,
we
provide.
D
We
provide
a
default
support
for
the
even
bus
which
is
like
you
can
define
a
negative,
a
nest
native
type,
even
even
bus,
and
then
we
will
create
a
this.
Even
bus
con
control
will
create
a
not
streaming
stepper
set
for
you
in
your
namespace
or
you
can
use
existing
even
but
not
streaming
and
then
created
even
bus
object
pointing
to
the
existing
live
streaming
service
and
we
are
open
to
implement
other
kind
of
email,
bus
like
kafka
or
what
any
kind
of
stuff
you're
interested
in.
D
So
this
is
our
architecture.
D
Some
of
the
interesting
features
provided
by
our
event,
so
first
one
is:
you
can
use
parameters.
So
when
you
receive
event
and
you
trigger
action-
you
actually
you
actually
want
to.
You
sometimes
want
to
customize
your
action
based
on
the
data
in
the
events
content.
D
We
call
this
feature
as
parameters,
so
you
can
just
customize
this
by
watch
the
events,
content
and
the
event
data
structure
within
the
system.
Whatever
it
is.
No,
it's
a
matter.
It's
a
it's
a
sqs
or
web
book
or
what
kind
of
or
any
kind
of
events
you'll
be
wrapped.
As
a
data
structure
like
this
you'll
see
it's
like
a
cloud
events,
type
format
and
the
parameter
will
you
can
read
the
event
context
talk
context
you
can
get
this
kind
of
property
in
your
action
or
any
kind
of
data.
D
From
your
original
immense
content
to
the
action
you
can
set
the
default
value
for
for
the
parameter
I
use.
You
can
also
use
like
tools
like
sprig
template,
to
make
some
conversion
for
your
data
before
and
trigger
the
action
and
operations
like
that
you
can.
So,
for
example,
you
have
a
you,
have
data
from
subject:
okay
and
then
you
want
to
add
this
subject
to
an
existing
value
or
just
replacing
replace
existing
value.
So
that's
the
operation.
D
D
D
D
What
actually
want
to
take,
but
sometimes
you
want
to
say,
hey
action,
one
I
only
want
one
actually
want
to
be
triggered
when
kind
of
dependency
or
kind
of
condition
happens
or
for
for
the
other
triggers.
I
want
to
use
a
different
conditions
like
this,
so
this
guy
I
wanted
this
trigger.
One
happens
when
only
when
the
dependencies
year
two
happen,
and
for
this
guy
I
wanna,
I
want
the
trigger
three
happens
when
this
condition
is
resolved
to
true.
D
D
Next
trigger
source,
so
arguments
can
be
used
to
trigger
any
kinds
of
kubernetes
resource,
which
is
any
kind
of
discriminant
object.
You
can
put
it
in
inside.
The
sensor
object.
The
raw
stack
in
the
sensor
object.
So
we
we,
we
know
it
has
inline
or
you
can
put
the
spec
in
a
config
map
or
github
or
a
simple
url
or
phone
or
three
bucket.
So
it's
kind
of
flexible
for
you
to
define
your
action,
and
now
I'm
going
to
give
some
demo
here.
D
D
So
I
here
I
have
a
cluster:
it's
a
k30
cluster
on
my
local
and
and
I'm
gonna
demo
like
three
or
four
very
quick,
real,
quick
once
okay,
so
first
I'm
gonna
give
a
hello
world.
D
Everything
starts
from
her
work,
so
this
hollow
word.
I
have
at
event
stores
I
use
a
calendar
for
a
demo
but
think
about
it
because
there's
a
category
cluster,
I
cannot
do
those
complex
external
independence
like
fqs
stuff,
so
I'm
going
to
use
calendar
or
web
hook
for
my
demos,
but
you
can
you
actually
can
change
any
like
change,
this
kind
of
even
sort
to
any
kind
of
resource
I
mentioned
in
those
25
and
then
the
result
is
the
same.
D
D
So
when
there's
an
event
happens,
I
want
to
trigger.
I
want
to
create
text
text
which
print
out
the
hello
world
all
right,
so
I'm
going
to
apply
this
spec.
We
actually
have
a
ui
for,
but
where
going
to
use
a
terminal.
D
Oh,
I
forgot
something,
so
we
have
a
one
bus
object,
so
we
need
to.
We
need
to
create
an
even
bus
object.
Direct
on
before
everything
happens,.
D
Test,
okay
and
then
after
I
create
a
humanbus
object
which
will
create
it's
right
now,
I'm
using
a
native
type
nas
and
this
crd
object
and
they
will
create
a
not
streaming
service
as
a
stable
set
in
my
namespace.
So
I
have
it
there
and
then
I
have
this.
Let
me
request
those
guys.
D
It's
still
same
task
wrong,
but
in
echo
the
hello
world,
I
changed
the
echo
hello
paris
message:
this
is
a
pure
tactile
spec,
so
here
in
arguments
I
definitely
define
a
parameter
here.
I
want
to
use
a
body
dom
message
from
dependency,
one
which
is
a
which
is
a
web
hook.
Example
here,
which
I
define
to
replace
to
replace
spec
parameters,
spec
parameter.
The
first
item
replace
the
value
which
is
the
word
okay.
D
D
Let's
express
my
8080
right,
okay
and
I'm
using
another
window
to
to
send
the
http
request.
I
already
have
it
there.
Okay,
so
make
it
quick.
So
I
sent
a
post
a
request
to
8080
with
a
message
demo.
So
the
task
one
should
print
out
a
message
like
a
hello
demo.
Instead
of
hello
world.
Let's
see,
okay,
all
right!
It's
a
new
task
around
here.
There.
D
It's
still
it's
similar
to
the
second
one,
but
I'm
adding
something
to
to
the
to
the
sensor
config
here.
So
it's
test
scoring
that's
still
same
by
using
the
parameter
stuff.
I
just
added
a
filter
to
the
dependency
here.
I
only
I'm
only
interesting
the
source
message,
which
has
a
body
dot
value
greater
than
50..
D
D
We
can
see
in
the
in
the
namespace
and
if
I
make
a
hp
request
to
the
endpoint,
I
already
have
it
here.
So
first
first
message
I
want
to
send
is
like
body.message
just
30.,
so
you
should
not
trigger
any
action
in
the
sensor
here.
We
don't
see
any
new
tasks
around
there,
but
if
I
change
it
to
like
60
all
right
all
right,
there's
new
task
room,
it
should
still
be
a
demo.
A
D
My
last
demo
will
be
a
trigger
condition,
so
for
this
demo
I
I'm
going
to
create
a
web
given
source.
It
has
two
webpage
events
defined
there.
First,
one
named
full,
it
listens,
part
8080
and
then
endpoint,
four
and
post
second
one
negative
bar
edit
one
bar
search
bar.
D
That's
a
given
source
card
and
for
sensor
I
have
I
defined
two
dependencies
there
and
with
bar
so
far
sensor.
You
have
to
define
the
dependence
from
different
event
source.
It
doesn't
have
to
be
from
the
same
event
source
and
then
I
have
two
triggers
here.
Two
test
runs
first
test
run.
It
has
a
condition
say:
hey,
either
full
or
bar.
D
When
I
see
a
message
from
either
full
or
bar,
I'm
gonna
create
a
task
run
and
general
name
is
a
general
name,
full
or
bar,
and
also
echo
pool
bar.
The
second
trigger
is
the
full
n
bar
that's
a
condition,
and
then
I
will
create
a
task
for
name
the
full
n-bar,
with
the
name
and
also
it
echoes,
like
full
n
bar.
So
that's
a
multiple
condition
example
and
I'm
gonna
grade
it.
D
So
we
can,
we
also
can
see
those
the
event
source
has
a
full
event
for
events
as
a
condition
for
a
or
bar
and
for
mbar,
in
that
that's
a
trigger
this
ui
used
to
this
line
used
to
blink
when
there's
events
coming,
but
I
don't
know,
definitely
is
there
something
wrong
with
it
to
fix,
so
this
line
is
not
blinking,
it's
just
always
right
out,
but
anyway,
so
now,
I'm
gonna.
So
because
I'm
exposing
88.81.
D
D
D
D
And
then
I'm
gonna
send
messages
to
those
two
endpoints.
So
before
that
I'm
check
I'm
checking
if
there
any
tasks
around
there.
Nope
and
I'm
gonna
send
a
first
message
to
full,
because
I
you
know
I
I
didn't
use
any
parameter
from
the
source
I
just
so.
I
just
send
it
amputate
json
screen
to
the
endpoint,
so
I
send
it
and
send
it
to
full
and
then
let's
see
full
or
bar.
So
that's
the
first
action
right.
I
can
see
there's
a
task
right
there.
D
D
B
On
the
the
and
condition,
how
long
does
it
wait
for
the
and
condition
to
become
true,
so,
in
other
words,
how
long
is
the
event
on
the
event
bus.
D
B
D
Got
it
yeah?
I
think
that
ends
on
my
demos
and
back
to
actually
I
don't
have
anything
more
question
actually
actually
so
for
even
for
arc
events
or
this
event
source
and
center
part
there,
you
can
do
hvac
for
them
everything
you
specify
replica
for
the
human
source
or
the
sense
sensor,
cr
object
and
then
we'll
do
active,
active
or
active
passive
hk
for
based
on
different
kind
of
event
source-
and
you
know
sensor.
This
processor
is
always
active
passive
for
event
source.
D
It
depends
if
it's
github,
github
or
those
kind
of
webhook
based.
You
could
do
an
active
active
for
like
calendar.
You
know
you
cannot
run
two
counters
at
the
same
time.
Other
you'll
get
duplicate
message.
So
so
we
run
active
passive
yeah.
That's
h,
a
yeah,
any
other
questions.
For
you
know.
I
don't
think
I
have
anything
else
to
to
share
today
the
free
to
answer.
Question.
E
One
question
on:
on
the
last
thing:
we
looked
at
the
triggering
on
on
events
of
different
types.
I'm
wondering
is
there
a
way
to
say
yeah,
I
want
to
trigger
one
there's
an
event
of
type
a
and
an
event
of
type
b,
but
that
share
some
property
like,
for
instance,
I
want
to
trigger
when
service
x
has
been
deployed
in
both
environments,
z
and
environment
y.
So
you,
like
declared
that
I
don't
know
what
this
shared
key
will
be,
but
they
will
have
the
same
value
of
some
sort.
E
I'm
thinking
about
a
fan
in
scenario
where
some
part
of
a
pipeline
starts
to
parallel
tasks
and
those
two
parallel
tasks
share
like
the
instance
idea,
run
id
or
something,
and
then
when
you
want
to
fan
in,
you
need
them
to
have
the
same
run
id
to
trigger
the
next
job.
I'm
not
sure
if
that
question
makes
sense,
but
that
is
a
scenario
that
I've
seen
before
at
least.
D
You
I
mean
you
still
can
use,
use
the
data.
I
mean
the
parameter
from
the
event
content
to
trigger
the
actions.
Basically,
that
can
you
can
like
instant,
instant
id,
but
actually
I'm
thinking
one
task.
You
have
two
one
one
pipeline
have
two
attacks
around,
but
that
you
can
define
two
event
source
for
them.
E
Yeah,
I
guess
a
similar
thing
would
be,
for
instance,
in
captain
there's
something
called
a
context.
So
whenever
captain
runs,
it
shares
the
same
context
key
across
everything
that
it
runs.
That
belongs
to
the
same
run,
so
to
say
so
I
might
want
to
use
argo
events
to
say
that
when
these
two
messages
come
or
these
two
events
come
and
they
belong
to
the
same
run
so
to
say,
then
I
want
the
trigger,
because
otherwise.
B
E
D
So
you're
saying
your
event:
source
actually
has
some
extra
conditions.
I
mean
it's
not
pure
a
and
b,
but
it's
like
something
some
condition
of
a
happens
and
some
condition
of
b
happens.
It
trigger
reaction.
E
C
Okay,
yeah,
I
hope
I'm
very
grateful
for
your
being
here
eric
and
that
it's
been
very
interesting
to
listen
to
you
and
I
actually
thought
if
we
should
interrupt
you
earlier,
but
because
we
have
more
things
to
discuss.
But
I
think
it
was
interesting.
So
I
wanted
you
to
to
finish
the
game
and
everything
and
it's
great
yeah.
D
C
Sure
and
you're,
of
course,
welcome
to
join
us
online
as
well
when
discussing
similar
use
cases,
and
I
think
the
use
case
that
eric
has
mentioned
is
something
that
we
might
come
back
to
in
a
detailed
discussion
on
the
vocabulary
and
how
to
set
up
this
system.
I
guess
maybe
our
pork,
the
the
new
version
of
the
pork,
could
maybe
maybe
involve
argo
events
we'll
see
about
that.
That
could
be
interesting,
actually
yeah,
I'm
sorry
to
take
too
much
time.
It's
it's
fine!
B
I
I
think
for
me
the
only
thing
that
we
probably
would
need
to
dive
into
is
the
the
since
we're
talking
about
the
format
of
of
events,
how
to
structure
them
kind
of
on
the
vocabulary
lines
that
that
would
come
into
play
down
the
road
down
the
road.
So
I
think
we
may
try
to
bring
you
back
when
we
get
deeper
into
the
the
weeds
on
our
our
our
event,
structures.
C
Yep
great
so
thanks
again
then
moving
on
then
to
our
next
topic
in
the
agenda.
That's
the
protocol
naming
andrea,
would
like
to
take
over.
A
Yes,
thanks
emil,
so
we're
trying
to
kind
of
make
some
progress
in
the
protocol
naming
topic
and
we
thought
with
the
ml.
The
first
thing
we
should
try
to
agree
on
is
what
what
are
the
requirement
for
a
good
name-
and
we
discussed
this
before
we
have
a
beat
up
discussion
as
well.
A
A
I
think
it
makes
sense
to
to
have
a
name
that
is
in
some
way
pronounceable
and
also
one
requirement
that
is
kind
of
hard
to
check
online
now,
but
we
can
look
into
it
as
the
name
is
available,
meaning
there
are
no
obvious
overlaps
in
terms
of
domain
being
used
already
or
yeah.
So
we
can
check
the
number
of
availability
in
terms
of
dns
domain.
For
instance,
also
things
like
github
organization,
twitter
accounts
and
so
forth,
other
requirements
or
negative
requirements.
A
If
you
will,
we
don't
want
the
name
to
bind
to
any
organization
or
group,
and
we
don't
want
the
name
to
bind
us
to
any
underlying
transformer
protocol,
because
these
are
things
that
might
change
so
the
protocol
that
we
are
defining
now
is
created
now
as
part
of
the
event
seek,
but
it
we,
we
hope
it
to
become
like
a
project
on
its
own,
probably
hosted
by
the
cdf.
A
A
Other
things
that
were
mentioned
it
would
be
nice
for
the
name
to
provide
a
nice
brand
name
to
something
that
may
be
nice
as
a
brand.
A
A
Another
comment
was
it
this
felt
as
it
sounds,
maybe
nice
to
have
it's
easier
to
it's
easy
to
dispel
it
and
something
else
we
mentioned
is
we
might
want
to
line
into
something
that
people
already
knows
so
there's
some
name
that
is
yeah,
I
guess
so.
People
can
associate
the
things
that
are
really
familiar
with
yeah,
the
last
four
that
didn't
consider
them
as
a
requirement
more
like
comments
or
nice
things
to
add
to
be
associated
with
a
name.
A
So
any
you
want
to
to
ask
if
there
is
any
any
comment
on
these.
If
this
makes
sense,
we
can
use
this
as
a
framework
of
reference
from
choosing
the.
B
F
Does
that
mean
that
we
should
not
be
discouraged
like
names
that
does
not
have
anything
with
the
thing
to
do
like
simeo
or
I
don't
know,
maybe
eve
was
one
other
name.
What
did
that
misunderstand
there.
E
I
think
tracy
had
a
comment
that
we're
we
could
use
two
different
names,
one
for
the
the
more
like,
not
gonna,
say
boring
parts,
but
the
more
like
strict
standard,
stuff
parts
and
another
for
for
like
project
where
the
projects
like
the
popkins
and
stuff
like
that
could
use
more
fancy
names.
But
the
the
more
boring
event
standard
stuff
would
be
more
related
to
the
topics.
I
think
that's.
What
was
what
tracy
brought
up.
G
G
You
know
at
the
best
I
could
what
we
were
trying
to
accomplish
and
I
felt
like
the
as
the
boring
part.
The
best
name
is
cd
events.
I
know
it
gives
us
it
assigns
us
to
that
part
of
it
assigns
us
to
the
continuous
delivery
foundation.
It
gives
us
the
continuous
delivery.
The
cd
part,
I
just
don't
know
how
quickly
cd's
going
to
go
away
and
it
matches
more
like
with
cloud
events,
and
I
think
people
would
get
what
we
were
doing
if
we
just
keep
calling
that
kind
of
that
standard.
G
The
cd
events,
but
if
we
have
a
project,
if
we
have
a
product
that
we're
going
to
be
like,
like
a
listener,
I
felt
it
should
have
a
different
kind
of
a
name
and
one
that's
more
relatable.
So
I
played
with
that
quite
a
bit
too.
I
try
to
create
an
acronym
and
then
I
wrote
down
evie
for
events,
and
I
was
like
oh,
we
could
just
call
it
ev
and
then
I
was
like
no.
G
B
G
C
Yeah
and
maybe
matthias
to
comment
on
your
question
also
there
but
related
to
the
topic,
and
you
mentioned
samuel.
That
was
one
of
my
suggestions
earlier.
We
will
discuss
later
on
if
the
current
suggestions
that
we
have,
if
they
do
align
to
these
requirements
or
not,
maybe
you
could
argue
that
cmu
is
still
something
it's
not
naturally,
aligned
to
the
first
first
requirement
there,
but
if
you
would
know
greek,
it's
kind
of
related,
but
we
can't
come
to
that
and.
C
B
I
think,
based
on
what
we
just
saw
from
derek
on
the
argo
side,
you
know
argo
ended
up
being
split
into
four
separate
pieces.
You
know
that
they
had
the
workflow,
you
know,
so
I
think
the
the
high
level
of
just
doing
the
the
cd
events
is
a
good
bucket
to
hold
the
other
pieces
that
will
come
out
of
that
in
place.
So
I
can
see
the
the
multiple
names
working
well.
A
Just
let
me
to
the
yeah,
so,
regarding
the
first
requirement,
I
agree
with
what
has
been
said
so
far.
I
think
the
the
fact
that
it
should
be
related
to
our
topic.
A
It's
it
doesn't
mean,
I
think,
like
names
like
simias
or
eevee
or
other
names
that
are
are
related
because
they
are
inspired
by
anyways
by
events
and
messages
and
so
forth.
So
it
didn't
wasn't
meant
to
be
saying
it
should
be
a
boring
name.
It
was
meant
to
say
they
should
be
related.
Somehow
that
was
my
interpretation
for
the
first
requirement,
so
it
shouldn't
be
something
that
is
totally
unrelated.
I
don't
know
name
about
food
or
something
that
is
not
related
at
all
with
messages.
A
So
thanks
and
tracy
and
steve
and
everyone
for
the
comments
on
these
I
mean
we.
We
had
a
set
of
proposed
names
until
now,
and
I
think
focusing
first
on,
let's
say
the
boring
name,
or
at
least
the
name
for
what
we
call
the
name
for
the
specification
of
the
protocol.
A
We
can
review
the
existing
list
just
wanted
to
show
it
briefly
here.
So.
A
Were
proposed
until
now,
and
many
of
them
are
indeed
somehow
related
to
other
city
events
or
cde
or
variations
of
it,
the
only
exceptions
being,
I
guess:
semio
simeos,
cse,
cdfe,
an
open
life
cycle
administration
leap
and
spell
so
in
terms
of
matching
with
the
requirements.
We
are
running
a
bit
short
in
time,
so
I
think
some
of
them,
like
cdfe
specific
to
the
continuous
delivery
foundation,
so
they
would
not
meet
the
requirements.
B
Just
just
do
a
strikethrough
on
that
so
on
on
the
ones
that
we
think
we
should
just
eliminate
immediately.
D
G
Because
I
think
that
the
the
orchestration
on
that
open
life
cycle
orchestration,
the
what
I
got
hooked
up
on-
was
orchestration
because
it
felt
more
like
jenkins
or
a
circle
ci
or
something
that's
doing
more
of
an
orchestration.
G
G
E
Just
a
quick
question
to
any
native
english
speakers
here:
cd
events
is
there
any
chance
that
a
malicious
actor
would
conflate
this
with
cd
like
the
cd
character
or
cd
bar,
or
something
like
that
and
use
that
against
us.
I
guess
it's
not
going
to
happen,
but
I
see
that
right.
I.
E
So
I
think
cd
means
like
in
disrepair
or
a
bit
sketchy
or
a
bit.
Yes,.
G
Nothing
wrong
with
just
calling
it
life
cycle
events
either.
If
we,
if
we
want
to
drop
cd
in
case
cd,
becomes
unpopular
in
the
future
for
some
reason
or
something
replaces
it
like,
you
know:
progressive
delivery,
life
cycle
events
is,
you
know
we
have
it
in
here
as
leap,
but
why
don't
we
just
call
it
life
cycle
events.
A
A
B
A
Yeah,
I
do
agree
on
that.
I
would
tend
to
take
the
ci1
out
unless
anyone-
okay,
I
I
do
like
the
the
idea
of
going
into
life
cycle
software
life
cycle
area.
A
I
just
my
only
concern
with
that-
is
that
it's
not
one
of
the
requirements,
but
it
might
align
less
to
things
that
people
expect
in
terms
of
cd.
C
C
Against
using
something
that
is
currently
the
best
practice,
so
I
mean
cd
is
what
we
talked
about
today,
but
we
might
talk
about
life
cycle
tomorrow
and
then
we
might
talk
about
software
supply
chain
in
the
even
after
tomorrow.
C
So
that's,
maybe
one
argument
for
not
having
any
of
those
so
say:
direct
related
words.
Let's
say.
G
I
would
on
that
I
would
argue
that
life
cycle
is
not
is
is
more
of
a.
How
can
I
phrase
it?
It's
more
of
a
concrete
word.
You
know
it's
it
it's.
I
don't
think.
Even
though
life
cycle
we
used
to
call
it
application
life
cycle
management
right
alm
before
we
started
calling
it
continuous
delivery,
but
life
cycle
really
does
describe
what
the
events
are
for,
as
opposed
to
a
trend
like
continuous
delivery.
To
me
is
a
is
more
of
a
trend,
kind
of
a
name.
G
B
A
Okay
right,
so,
I
think,
stronger
leaf,
stronger
candidates.
I
don't
think
we
have
an
abuse
winner
and
we
only
have
three
minutes
left,
but
I
think
we
we
are
like
strong.
Some
strong
candidates
like
cd
events
or
something
related
to
life
cycle
events,
and
I
think
names
like
simio
or
cemus-
are
still
an
option.
A
Yeah,
so
my
our
plan
was
okay.
If
he
had
like
an
obvious
wiener
to
send
an
email
with
saying:
okay,
we
propose
this
name
and
if
anyone
is
against
it,
please
let
us
know
at
the
certain
time
of
time
until
a
certain
point
of
time.
But
since
we
have
a
couple
more
names,
what
I
can
do,
I
can
build
a
new
poll,
a
new
poll
with
the
names
that
were
suggested
today.
A
So
we
just
do
a
free
options
out
of
these,
and
maybe
one
open-ended
option
if
people
want
to
add-
and
we
can
see
if
there
is
a
strong
winner-
and
hopefully
we
can
close
this
and
pick
a
name
in
the
next
couple
of
weeks.
Does
this
make
sense
as
a
plan.
G
A
G
Do
want
to
say
that
I
liked
simeo,
but
I
looked
that
up
and
there
are
there's
some
startups
that
have
already
started
using
that.
G
C
I
think
also
before
we
settle
on
the
name,
we
should
look
it
up
if
it's
already
used
somewhere
and
if
the
domains
and
such
are
made
available
before
we
act
at
least
set
on
an
abbreviation
of
it,
whatever
we
should
use
so
that
leave
that
are
out,
there
might
be
something
that
we
could
abbreviate
lifecycle
events
to,
if
that's
not
occupied,
so
maybe
the
abbreviation
we
might
not
need
to
decide
on
exactly
now
or
to
the
poll,
but
at
least,
if
it's
okay
to
use
life
cycle
or
not
in
the
name,
that's
something
we
should
try
to
distinguish
them.
G
So
I
just
I'm
just
looking
now:
there
are
some
references
that
people
are
using
the
term
life
cycle
events,
but
I
don't
see
anything
that
is
called
it.
G
A
All
right,
okay,
yeah.
I
think
we
are
very
client
but
I'll
make
this
poll,
and
what
I
can
do
is
also
reach
out
to
the
cdf
and
see
if
there
is
any
service
that
can
maybe
provide
us
in
terms
of
branding
check
and
make
sure
there
is
no
like
trademark
infringements
or
anything
like
that.
And
when
we
pick
a
name
and
yeah.
B
A
I'll
try
I'll
send
out
the
poll
as
soon
as
as
possible.
Then
great.