►
From YouTube: CDF SIG Events - May 8, 2023
Description
For more Continuous Delivery Foundation content, check out our blog: https://cd.foundation/blog/
B
C
I
can
I
can
take
you
folks
through
what
we
discussed
last
time.
We
didn't
get
through
everything
in
the
reference
architecture,
deck
that
we
put
together.
B
B
Let's
see
yeah
I,
don't
expect.
We
will
have
more
participants
now.
So,
let's,
let's
kick
off
thanks
Steve
for
letting
us
as
well
with
your
name
there
today
anything's
minutes
so
just
quickly.
Before
we
get
to
the
bsmi
CC
the
interoperability
parts,
we
can
see
some
updates
from
what
happens
around
us.
B
So
I
said
last
time:
I
should
prepare
some
some
materials
that
should
be
presented
to
the
technical
oversight,
Committee
of
CDF,
recording
the
tig
events
and
the
project,
but
so
far
I
haven't
prepared
that
that
remains
to
be
done.
Andrea
had
an
action
to
update
the
cdnance
white
paper
for
for
the
Silicon
now,
and
that
is
done.
I
think
that
the
white
paper
will
be
published
this
week.
Yes,
very
small
updates
to
the
old
one,
some
quick
mentioning
of
what
happens
in
the
city
dance
work
group.
B
We
released
version
0.3
last
week
on
the
series
events
protocol
that
contains
testing
events
that
were
contributed
from
the
test,
Cube
people
and
also
signing
artifacts
a
new
event
for
for
declaring
that
artifact
has
been
signed.
B
B
There
will
be
some
elections
coming
up
of
project
representatives
in
the
TOC
I
just
want
to
mention
it
briefly
to
you
know.
So
we
have
it
recorded.
There
will
be
four
seats
in
the
TOC
available.
They
will
be
elected
every
second
year,
I
think
it
was,
and
then
those
four
seats
will
be
it'll
be
filled
with
people
from
the
nine
different
projects
that
we
have
and
apparently
then
not
all
projects
will
have
a
representative
on
the
TOC.
B
So
what
we
need
to
do
not
now,
but
between
May,
22nd
and
June
5th.
We
should
provide
one
name
from
this
project
that
should
run
for
elections
towards
the
TOC,
so
we
will
come
back
to
that.
That
should
be
as
well
I've
seen
forming
about
the
process.
There
then
quickly,
we
had
a
meet
up
last
week
as
well
with
the
cncf
group,
the
cncf
tag,
delivery,
I
think
it's
actually
case.
B
It's
a
technical
advice,
regroup
I
think
it's
called
and
we
discussed
a
bit
from
different
ways
to
collaborate
there
as
well,
and
one
idea
that
came
up
was
that
we
should
Implement
City
events,
support
in
the
Potato
Head
implementation,
that
many
people
in
the
tag
up,
delivery
or
or
involved
in,
and
we
also
said
that.
Yes,
this
is
an
interesting
collaboration.
We
should
continue
and
we
will
meet
again
in
like
a
month
or
so
the
white
paper
I
mentioned
it's
already.
It's
probably
to
be
published
this
week.
C
B
Might
have
come
from
me,
it
might
have
come
from
you,
okay,
yeah
and
I
just
wanted
to
note
that
we
don't
have
anyone
using
it
in
production
at
the
moment.
Sincere
event
is
not
released
in
a
in
a
1.0
version.
Well,
of
course
it
could
have
been
used
anyway
in
production,
but
no
one
like
it's
just
there
to
do
that
at
least
at
this
moment.
B
C
Have
there
been
any
vendors
signed
on
to
adopt
CD
events
or
intend
to
adopt
CD
events.
B
A
I,
don't
know
actually
so
like
having
that
RFC
approved
for
Spinnaker,
and
then
we
are
doing
the
integration
of
CD
events
to
Spinnaker.
So
will
that
be
kind
of
official
agreement
I'm
not
sure
like
if
it
is
anything
needs
to
be
said
officially
like
we
are
saying,
Spinnaker
integration
with
CD,
then
right.
B
To
implement
that
so,
as
I
mean
open
source
organizations
such
as
Jenkins
and
Spinnaker,
yes,
they
will.
They
will
give
you
and
support,
but
maybe
not
Enterprise,
vendors
or
cicd
tools,
I
think
that
was
I
I
interpreted.
That
was
what
you
said:
yeah
personalizations
for
sure
they
will.
They
aim
to
support
syrians
many
of
them.
B
Okay,
and
that
quick
introduction
leads
us
now
to
the
vs
my
discussion
unless,
for
example,
usually
under
just
if
you
want
to
have
something
else
set
before
we
go
into
the
vs.
My
discussion.
D
A
So
the
thing
only
one
thing
I
wanted
to
check
here.
Actually
so,
if
you
wanted
to
so,
there
are
like
different
types
of
CD
events
right
so
let's
say
in
Spinnaker
Point
of
View
I'm
talking
about
because
currently
is
what
I'm
trying
to
implement
within
the
Spinnaker
so
see
like
Spinnaker
is
able
to
produce
some
CD
wins
right.
So
what
in
case
of
like
one
pipeline
runs
finished
or
something
happens
like
Spinnaker
decides
to
send
some
events
out
of
the
Spinnaker
to
events
broker
yeah.
D
A
Is
there
any
thing
that
we
need
to
restrict
on
which
type
of
CD
events
can
Spinnaker
send
are
like?
We
should
be
giving
a
capability
of
sending
all
type
of
events
from
Spinnaker
to
the
outside
wall.
So
because,
when
asking
is
like
some
events
of
CD
event,
types
may
not
be
related
to
the
Spinnaker
case,
I'm,
not
sure
like.
In
that
case,
let's
say
the
source
code
Version
Control
related.
D
A
Will
it
will
it
be
like,
irrespective
of
anything
like
will
not
bother
about
like
what
event
you
send?
You
send
it
we'll
say
Spinnaker
that
will
give
like
the
support
for
all
CD,
even
types
to
send
Prince
back
around.
So
it's.
D
B
That
I
I
think
that's
well,
it
depends
I
guess,
but
but
I
think
that
Spinnaker
and
another
such
orchestration
tools,
if
we
should
group
them
as
that
such
they
should
probably
send.
Events
like
the
pipeline
started,
finished
events
and
the
task
run
finished
and
started
events
forever
appropriate.
So
when,
when
some
execution
is
happening
in
those
tools,
they
should
probably
send
those
events,
if
you
configure
them
to
do
so,
of
course,
that
shouldn't
be
mandatory,
I
would
say,
but
I
at
the
same
time,
I
think
it
should
be
implicitly
sent
in
the
background.
B
But
then
it
could
also
be
that
spinach,
for
example,
has
support
for
additional
events.
Like
you
said,
deployed
events,
the
ultra
deployments,
whether
that
should
be
implicit
or
it
should
be
just
supported
by
Spinnaker,
so
that
the
users
can
explicitly
call
it
in
some
in
some
step
or
their
Spinnaker
pipeline
I.
Think
that
adds
up
more
or
less
for
the
Spinnaker
Community
to
decide,
and
at
this
moment
I
mean
City.
Events
doesn't
put
any
requirements
on
such
tools
that
they
need
to
be
able
to
send
a
certain
group
of
events.
B
Maybe
we
will
have
such
lists
eventually,
but
not
not
at
this
moment.
So
the
the
main
thing
we're
saying
that
at
all
supports
City
events,
is
that
it's
able
to
to
send
at
least
some
kinds
of
events
and
it's
configurable
how
to
well.
The
content
of
them
should
somehow
be
configurable.
I
would
say.
D
B
So
Steve,
let's
go
into
the
vs
my
interoperability
and
did
you
have
something
in
mind
to
to
continue
the
discussion
based
on
or
I?
Do.
C
Okay,
so
there's
this
actually
should
be
shared
with
you.
C
Folks
I
think
it
should
be
open,
but
I'll
send
you
the
link,
just
in
case
my
screen
share
is
too
small
or
something
so
currently
we
have
a
lot
of
focus
on
this
kind
of
delivery
mechanisms,
pipelines
and
platforms,
but
not
necessarily
the
the
auditability
or
management
of
the
workflow
that's
happening
inside
of
those
and
on
top
of
those
platforms,
so
value
stream
management
platforms
really
kind
of
strive
to
act
as
a
layer,
above
all
the
tooling,
that
is
contributing
to
this
ultimate
value:
delivery,
right,
the
entire
end-to-end
workflow,
so
broader
scope,
but
also
like
a
higher
level
above
individual
Point,
Solutions
or
even
Suites,
because
we
have
different
domains
that
come
together
in
a
value
stream.
C
It's
very,
very
Broad,
so
there's
limited
visibility
because
it's
siled
in
different
systems.
It's
very
difficult
to
attribute
effort
to
outcome,
so
we
don't
really
understand
what
impact
any
specific
practices
are
having
or
how
work
is
actually
able
to
contribute
to
outcomes
and
things
like
access
to
data
and
the
format
of
data
really
get
in
the
way
of
adopting
any
of
these
Solutions.
C
Whether
it's
commercial
or
open
source,
a
lot
of
the
commercial
vendors,
have
solved
this
problem
separately,
but
there's
very
limited
possibility
in
the
open
source
world
because
we're
dealing
with
many
different
apis
and
sdks
and
data
in
different
formats
when
we're
typically
just
talking
about
the
same
thing
right.
We're
really
just
talking
about
the
flow
of
work
across
abroad,
Spectrum
or
a
an
extended
workflow.
Let's
say
yeah.
D
C
Go
through
the
link
to
catch
up,
that's
good,
perfect!
So,
essentially,
where
we
got
to
last
time.
We
reviewed
up
into
like
the
similarities
and
contrasts
with
CD
events.
C
So
fundamentally,
there
seems
to
be
a
lot
in
common,
but
but
there
are
some
notable
differences
right,
we're
talking
about
a
broader
scope
from
Beyond
cicd,
all
the
way
to
like
starting
the
the
workflow
all
the
way
to
completing
a
workflow
and
then
I
think
we're
primarily
dealing
with
the
different
substrate.
A
different
data
store,
I
think
that
that
depends
I'm,
not
quite
sure,
where
you're,
where
you're
always
or
consistently
getting
data
from
and
then
in
the
CD
events
world.
C
C
Management
of
dependency
is
this
infrastructure
code
all
the
way
through
to
actually
I'm
kind
of
lost
on
some
of
these
I
forget
I
think
this
is
actually
your
phone
all
right,
yeah,
grafana
right
so,
but
essentially
we're
looking
at
tracking
work
all
the
way
through
all
these
systems,
the
same
workflow
or
the
same
work
items
actually
I
should
say
you
know
we
have
something
that
we
want
to
do
and
it
gets
decomposed
and
clarified
and
refined
and
built
and
tested
and
delivered
and
operated
and
analyzed
to
understand
whether
the
thing
that
we
wanted
to
do
actually
turned
out
to
be
everything
we
hoped
it
was
and
across
that
landscape.
C
We
have
a
number
of
these
different
tool
categories
that
are
represented
across
the
value
stream
and
then
some
broader
areas
below
that
are
kind
of.
C
C
It
needs
to
comply
with
governance
and
Regulatory
obligations
and
support
quality
management,
but
there's
this
broad
range
of
tooling
that
goes
beyond
your
typical
CI
CD
scope,
and
that
being
said,
the
commonalities
with
you
know,
you
know
time,
series,
data
and
event
driven
data
I
think
are
pretty
undeniable,
so
there's
a
lot
of
similarities
with
something
like
CD
event,
scope
and,
let's
say
methodology.
C
So
hopefully
that
means
that
there
is
something
that
we
can
borrow
and
something
that
we
can
also
contribute
to
in
return,
but
I'll
I'll
leave
it
there
and
pause
for
any
comments
or
questions.
B
Okay,
yeah
I
I
I
do
believe
that
there
are
a
lot
of
things
going
on
in
parallel
in
this
flows,
I
mean
the
crcd
pipeline
is
one
thing
that
happens,
but
in
parallel
to
that,
as
you
say,
also
I
mean
Financial
systems
might
be
working
and
well
Information,
Management,
product
handling
or
some
kind
of
PLM
or
Alm
processes
as
well.
That
are
managed
at
the
same
time,
which
might
not
be
explicitly
part
of
the
cicd
set
up
for
the
cicd
pipelines.
B
If
we
say
so,
and
in
some
way
they
need
to
be
interconnected,
I
believe
so
that
you
can
trace
changes
in
different
areas
and
their
effects
back
and
forth
between
these
different
area.
So
what
we
should
call
them.
So
we
have
there
to
the
right.
I
mean
continue.
Integration,
delivery
and
Source
control
management
and
I.
B
C
Expectation
is
that
the
lowest
barrier
to
entry
for
this
data
would
be
a
let's
say,
additional
data
source
that
compiles
data
from
all
these
existing
data
stores
that
we
already
have
right
I
mean
we
already
have
a
lot
of
the
data
that
is
necessary
to
address
the
script
in
these
separate
systems.
But
it's
you
know
those
systems
have
more
data
than
we
need,
and
it's
not
formatted
in
a
in
a
a
structure
that
is
conducive
to
value
stream.
Analysis,
they're
things
that
you
know
we
really
don't
care
about
in
the
Valley
Stream
world.
C
We
don't
care
about
individual
identities
as
much,
although
there
are
some
that
are
very,
very
interested
in
that,
but
it
doesn't
necessarily
pertain
to
the
performance
of
the
workflow.
Sorry.
C
Exactly
individual
contributors,
for
instance
people,
okay,
yeah,
yeah,
okay,
but
you
know
that
that
is
valuable
to
some
folks.
So
you
know:
I'm
envisioning,
there's
a
there's
a
big
opportunity
to
sort
of
pull
data
from
existing
data
stores
and
transform
it
into
something
that
is
conducive
to
Value
stream
analysis.
C
And
hopefully
you
know
at
a
level
that
folks
are
comfortable
with
right.
That
fits
the
governance
and
Regulatory
Compliance,
but
also
the
policies
and
utility
of
these
organizations
that
want
to
do
this
value
stream
analysis.
So
If
part
of
your
goal
is
to
understand.
Okay,
we
want
to
make
sure
that
individual
actions
by
individual
contributors
are
attributable
to
ultimate
customer
outcomes.
We
want
to
be
able
to
trace
okay.
That
person
did
that
thing,
and
it
was
part
of
this
thing
that
went
out
into
production,
which
I
think
is
not.
C
You
know
it's
not
my
primary
interest.
I,
don't
think
it's
a
very
compelling
use
case
for
Value
stream
analysis,
but
you
could.
You
could
perhaps
include
that
data
and
represented
in
a
in
the
context
of
a
value
stream,
but
primarily
the
the
most
important
thing
I
think
is,
is
to
be
able
to
say.
Okay,
we
have
all
of
our
jira
data,
backed
up
into
a
data
lake
or
data
warehouse.
C
We
have
servicenow
data,
we
have
CI
CD
data,
we
have
all
of
our
task
management
data,
they're
Loosely
correlated
on
their
own,
but
they
could
be.
C
B
Could
be
so
what
we
are
discussing
in
the
parallel
activity
on
how
to
connect
events
to
each
other?
How
to
correlate
events,
and
one
way
to
do
that
is,
as
you
say,
you
have
a
like
a
global
identifier
for
all
events
from
a
certain
flow
or
something
regroup
them.
But
then
you
would
also
probably
need
some
way
to
lay
them
out
in
time
and
to
be
able
to
see
like
a
traceability
graph
of
what's
happened
when
and
after
what
and
why
and
so
on.
B
So
it
might
not
be
enough
with
just
a
global
ID
and
therefore
we're
also
looking
into
the
concept
of
links
between
explicit
events.
So
one
event
could
have
some
kind
of
Link
reference
to
another
event
itself,
so
we
are
still
discussing
discussing
what
what
way
to
order,
but
I
I
am
I
was
thinking
what
we
have
been
discussing
a
bit.
Where
are
the
boundaries
for
cedarbon's
protocol?
B
What
kind
of
entities
should
we
deal
with
and
what
kind
of
parameters
of
those
entities
then
as
well,
and
it's
not
so
easily
defined
where
to
draw
the
line,
even
though
we
say
it's
cicd
related
events,
then
it
might
be
a
bit
obvious
in
some
discussions
at
least
well.
This
is
the
cicd
event.
This
is
not,
but
it's
it's
not
always
that
easy
and
I'm
thinking
now.
B
B
We
are
interested
in
industrial
entities,
for
example
when,
when
a
certain
developer
wants
to
follow
his
own
commit
from
push
to
to
deployment,
for
example,
be
able
to
to
filter
on
that
user
ID,
for
example,
and
that
makes
me
think
that
we
might
not
have
a
full
subset
I
mean
seriously.
It
might
not
just
be
a
subset
of
of
the
VSM
entities,
but
throughout
the
race.
C
C
What
is
the
path
of
any
one
individual
work
item,
and
it's
it's
same
with
same
with
the
kind
of
like
the
the
relation
between
groups
of
people
and
individual
people.
Right
I
mean
that
the
the
system,
the
larger
entity,
is
always
I,
think
more
representative
of
the
system
performance
than
any
individual.
C
There's
a
use
case
for
security
where,
in
security
they
want
to
be
able
to
trace
back
everything.
That's
one
individual
worked
on
back
to
Inception
and
that
could
be
anything
from
you
know
that
person
just
left
the
company.
So
we
want
to
make
sure
that
everything
that
they
worked
on
is
checked
or
validated,
or
you
know
we
need
to
hand
off
all
this
person's
work
to
this
person
because
they're
going
on
vacation
there's
any
number.
C
Not
I
wouldn't
say
directly,
no
but
I,
think
indirectly,
while
you're
while
you're
capturing
all
the
work
item
data
you
could
do
that
you
could
facilitate
that
capability.
But
no,
as
I
mentioned,
like
the
the
primary
focus
is
the
system
of
work
and
people.
Individual
contributors
are
kind
of
secondary
to
the
performance
of
the
whole
system,
yeah
system,
being
the
system
of
work,
including
all
the
people
and
practices
and
all
that
stuff
in
a
very
broad
sense.
B
So,
if
I
try
to
make
a
small
or
maybe
a
small
conclusion
here,
but
I
so
value
stream
management
is
about
seeing
how
value
is
added
or
changed
on
a
system
in
a
very,
very
simplistic
term,
while
cicd
is
maybe
more
low,
oriented
or
as
as
more
of
an
interest
to
see
individual
executions
of
certain
flow
or
certain
pipeline
that
we
call
it
so
they,
the
items,
the
the
main
items
of
say
in
the
crcd
system,
is
maybe
not
the
the
application
that
we
developed
or
the
the
system
that
is
evolving,
but
rather
pipelines
and
then,
of
course,
actions
in
those
and
many
artifacts
built
there
and,
of
course,
then
artifacts
built.
B
They
relate
to
the
system,
but
it
is
being
developed.
It's
not
the
main
concern
the
system
itself,
but
it
seems
like
value
stream.
Management
is
more
about
looking
at
the
system
and
how
does
the?
How
does
the
value
of
that
system
change
over
time
and
when,
when
it's
new
value,
added
and
so
on?
How
does
it
work?
Is
that
true.
C
C
Ultimately,
the
way
that
I
think
about
value
in
a
value
stream
management
context
and
the
difference
between
value
stream
management
and
something
like
Ci,
CD
or
or
even
devops,
is
that,
like
value,
is
really
just
explicitly
calling
out
the
fact
that
the
scope
in
question
is
all
the
way
from
the
Inception,
the
original
hypothesis
or
idea
of
we
sh.
We
should
do
this
thing.
It's
valuable.
We
think
it's
going
to
be
valuable,
all
the
way
to
the
reconciliation
of
that
with
a
customer
who
says
yes
that
was.
That
was
good.
C
Because
what
you're
seeing
sort
of
flow
through
this,
the
system
of
work
is
we're
trying
to
do
this
thing,
and
this
thing
is
it's:
maybe
it's
a
feature,
maybe
it's
a
bug
fix.
It
is
comprised
of
maybe
many
different
commits,
maybe
commits
to
many
different
systems.
C
There
could
be
multiple
branches
and
multiple
different
code
bases
that
are
a
part
of
that
yeah
and
so
we're
kind
of
like
a
level
of
abstraction
above
yeah.
What
I
think
is
commonly
captured
inside
of
cicd,
but
the
same
kind
of
activities
are
involved
just
like
at
a
different
level
of
atomicity
or
like
granularity,
yeah.
B
B
We
should
include
properties
of
the
the
events
that
occur
in
the
system
in
the
actual
events
that
are
sent
out
the
event
entities
and
in
some
cases
we
we
include
quite
a
lot
of
information
from
those
systems
and
in
some
cases
we
just
include
modest
references
to
us
that
a
specific
ID
in
those
systems
from
where
you
can
go
fetch
any
more
additional
data.
And
that
brings
us
back
a
bit
to
the
the
user
identity
discussion,
for
example.
B
B
So
so
then,
if
we
would
just
look
at
the
the
total
list
of
properties
in
the
in
the
events
that
we
send
out
in
the
City
events
themselves,
that
might
not
be
very
complete
picture
of
everything
that
is
involved
in
the
CRC
system.
But
you
will
need
to
go
fetch
data
from
elsewhere
and
maybe
some
such
data
might
be
needed
to
calculate
the
values
value
gains,
for
example
on
the
specific.
B
Thank
you
so
so
I'm
thinking
about
this,
when
we
should
store
data
somewhere
for
later.
C
But
I
think
that
that
is
true
also
for
the
value
stream
management
context
like
in
any
case
where
you're
pulling
data
on
work
items.
The
work
items
themselves
will
be
connected
to
all
kinds
of
other
attributes
right,
they're
part
of
a
overall
initiative.
They
had
specific
contributors,
they
involve
specific
code
bases.
C
All
those
things
could
be
retrieved
from
the
original
work
item
and
you
could
really
just
be
tracing
the
path
of
the
work
item
and
any
data
that's
relevant
to
understanding
the
performance
of
the
system
as
the
primary
goal
right
and
then,
if
you
wanted
to
understand,
you
know
who
was
involved
in
any
specific
aspect
of
the
workflow
you
could
go.
B
C
That's
another
thing
that
really
the
kind
of
the
the
identity
piece
was
introduced
by
a
bunch
of
folks
who
were
initially
involved
and
their
project
got
disrupted
and
I
was
working
with
a
company
that
was
kind
of
got
excited
about
this
and
they
had
a
bunch
of
folks
join
and
then
their
their
project
got
kind
of
dismantled.
C
But
the
anticipation
was
you
have
software
bill
of
materials?
Is
now
a
big
deal,
people
understanding
you
know:
where
did
this
software
come
from
right
words?
What
are
all
the
dependencies
included?
What
are
all
the
libraries
and
the
sources
of
those
libraries
in
the
provenance
of
all
those
things
and
who
contributed
to
creating
this
thing?
C
This
looking
forward,
you
know
10
years
into
the
future.
The
estimation
was
okay.
Well,
the
same
thing's
gonna
come
for
workflow
right,
where's,
the
workflow
bill
of
materials,
and
that
was
just
you
know
an
estimation
but
kind
of
like
I
think
a
safe
bet
that
at
some
point
people
are
going
to
want
to
know.
C
How
did
we
actually
end
up
with
this
thing
that
we're
putting
out
into
the
world
and
and
be
able
to
sort
of
like
sign
for
that
sign
for
that
workflow,
but
that
was
kind
of
the
motivation
of
okay?
Well
right
now
we
don't
know
who
is
necessarily
contributing
to
things
in
production.
You
can't
trace
something
that
was
that's
discovered
in
production
too,
who
worked
on
it
very
effectively,
but
I
would
say
that's
a
secondary
use
case
at
best,
like
the
primary
use
case
is
still
going
to
be.
C
B
No,
so
they
talk
about
this
bombs.
There
I
connect
that
creatively
to
data
needed
for
the
Alm
systems,
let's
say
or
the
maybe
software
PLM
system
to
handle
the
and
such
things
and
then
the
requirement
on
on
tracing
also
the
workflow
data
or
the
configurations
of
a
workflow.
Or
what
have
you
do
you
see
that
also
as
part
of
the
the
Alm
domain,
or
is
that
some
something
else
tracing
how
something
has
come
to
be
not
just
what
it
comprises
of.
C
C
Strongly
correlated
to
Value
stream
management,
value
stream
management,
I
think
is
basically
gaining
popularity
because
of
value,
and
so
the
the
desire
to
connect
efforts
to
value
and
understand
that
people
are
doing
ultimately
the
right
thing,
because
we
can
do
all
kinds
of
things
now,
but
then
they
also
the
stream
aspect
of
value
stream
is
very
appealing
to
folks.
People
are
very
into
flow
and
very
into
continuous
things,
so,
like
the.
C
Yeah
it
is,
it
is,
and
it's
funny
how
you
know
the
right
name
can
help
something
take
off,
but
I
think
that's
a
lot
of
the
momentum
and
I
think
it's
really
sitting
on
a
foundation
of
Alm
and
a
number
of
other
things
like
product
portfolio
management,
and
it's
really
just
because
it's
a
catcher
title,
but
it's
a
lot
of
the
same
stuff
and
with
and
I
will
say
that
with
a
slightly
different
Focus
right.
C
When
you
talk
about
application
life
cycle
management,
the
primary
unit
of
focus
is
the
application
which
you
know
any
individual
application
is
not
that
important.
They
change
over
time.
What
is
the
boundaries
of
the
application?
I
mean
like
the
application
is
comprised
many
different
things.
It's
different
things
to
different
people,
it's
less
important,
then
the
system
of
work
that
can
create
any
application
and
any
number
of
applications
and
allow
them
to
be
changed
over
time
and
ultimately
deliver
value
to
customers.
C
So
I
think
it's
really
just
a
refinement
of
the
the
mindset
around
what's
already
very
present
in
Alm
and
other
in
other
areas
of
focus.
Like
the
other
day,
I
drew
up
this
map
of
like
what
is
what
is
value
stream
management.
It's
like
50
different
things,
so
we've
had
we've
had
for
a
long
time.
You
know
everything
from
change
management
to
to
Agile
and
all
these
other
things.
B
C
It's
I,
don't
think
it's
common
in
the
Valley
Stream
management,
space
really
I,
I,
say
it
quite
a
bit,
but
I
don't
know
if
it's
a
if,
if
it's
like
would
be
considered
part
of
the
Lexicon.
B
C
C
Yeah
yeah
I
mean
it's
I.
Think
system
of
work
is
is
helpful
because
it
focuses
people
on
the
journey
rather
than
the
destination
right.
I
mean
a
lot
of
this.
A
lot
of
this
area
of
study
is
is
the
study
of
the
journey
that
delivers
you
to
the
destination
rather
than
focusing
too
much
on
the
destination,
or
you
know
a
a
piece
of
the
journey.
So
it's
really
just
like
how?
How
did
you
get
to?
C
The
journey
is,
is
a
tough
one,
because
everybody,
these
days
is
familiar
with
the
customer
Journey,
which
is
different
than
a
value
stream,
which
most
people
wouldn't
realize
unless
they've
had
to
kind
of
reconcile
the
two
things,
but
you
would
think
that
they're
very
similar
and
they
are
in
in
many
ways
right
I
mean
anything
that
sort
of
follows.
The
flow
of
time
can
look
very
similar
if
you
oversimplify,
but
something
like
a
customer
Journey.
C
C
You
know,
I
see
a
visual
of
like
a
customer
Journey
and
value
streams,
kind
of
feeding
the
motivators,
so
the
the
offerings,
the
services,
the
applications,
products
that
move
folks
through
and
one
of
those
is
you
know
it's
like
it's
as
simple
as
like
marketing-
is
one
value
stream,
where
they're
they're,
creating
and
delivering
campaigns
that
create
attractive
or
attention
grabbing
valuable
experiences
to
customers
in
a
very
broad
sense.
C
You
know
whatever
advertisements
or
engagement
or
something-
and
that's
you
know,
that's
where
we
get
into
the
scope-
that's
potentially
far
broader
than
something
like
cicd,
because
cicd
has
not
yet
made
it
into.
Hopefully,
at
some
point,
it
makes
it
into
marketing.
As
you
know,
my
core
substrate,
but.
C
C
Is
really
just
to
a
bunch
of
workflows,
interconnected
right.
D
B
Interesting
stuff,
so
where
are
we
going
next
now,
with
this.
C
Well,
it
seems
like
it
seems
like
we're
we're
both
fairly
early,
but
I
have
yet
to
I
mean
you're,
certainly
further
ahead
than
we
are
in
the.
B
Valley
in
sometimes
maybe
but
not
not
in
all
senses,
well,
I
might
be
from
my
background
of
another
events
protocol
the
Eiffel
event
protocol,
because
that
is
used
in
Productions
since
over
10
years
did
events.
Protocol
itself
is
very
new
and
it's,
as
we
said,
not
in
production.
Yet
so
that's
very
new
yet.
But
the
idea
of
propagating
data
through
events
from
a
series
system
is
not
at
all
new.
D
B
C
Yeah
I
wonder
yeah
I,
wonder
what
that
looks
like
myself
in
terms
of
like
the
foundation
of
something
like
City
events.
Well,
let's
say
that
that
the
data
store
aspect
of
CD
events
is
is
a
primarily
real
time.
Well,.
B
B
So
but
then,
of
course,
it
is
a
secondary
requirement
or
a
requirement.
I
would
say
on
a
system
that
is
City
events
aware
to
also
be
able
to
store
that
data
somewhere
and
be
able
to
retrieve
it
later.
But
that's
that's
more
of
the
on
the
ecosystem
around
the
city
events
protocol
and,
of
course,
if
you
implement
the
protocol
in
some
tool,
you
will
need
also
to
be
able
to
actually
emit
those
events.
You
will
need
some
kind
of
broker
or
bus
to
to
send
the
events
on
and
then
obviously
as
well.
B
You
need
to
store
it
somehow
somewhere
to
be
able
to
visualize
it
later
on
or
calculate
metrics
on
it
or
what
have
you
but
the
core
of
the
City
events
itself
is
just
an
event
protocol,
so
we
can
discuss
any
surrounding
system
as
well,
of
course,
which
is
part
of
the
whole
picture
where
events
is
used.
C
B
But
then,
if
we
talk
about
visualization
or
metrics
collections
or
anything
like
that,
then
of
course
it's.
We
also
need
to
store
that
data.
That's
close
in
those
interfaces
in
some
way,
but
it
depends
on
the
use
case.
So
if
you,
for
example,
we
talked
to
jalander
here
about
Spinnaker,
the
Spinnaker
is
just
just
interested
in
really
the
communication
itself
so
and
other
tools
like
gankings
and
other
such
orchestrators
are
probably
just
interested
in
being
able
to
trigger
on
events
and
send
out
to
them
others
to
trigger
on.
C
So,
in
that
case
you
know
I
think
there
might
be
there
might
be
somewhat
of
a
limited
overlap,
I'm
sure,
there's
quite
a
bit
borrow
and
maybe
something
you
know.
I
I
see
this
as
mostly
us
borrowing
from
what
you've
done
and
we
probably
don't
have
a
lot.
We
can
offer
you
in
terms
of
what
would
be
valuable
in
the
context
of
CD
events,
but
well
because
what
I
imagine
you
know
the
primary
operating
model
of
something
like
psmi.
C
C
I
think
it's
primarily
vocabulary.
I
think
consistent
vocabulary
is
very
valuable,
because
I
think
that
the
storage
aspect
I,
think
you
know
most
of
what
we're
we
actually
need
to
work
with
and
need
to
do
so
need
to
work
with
is
data
that
is
aggregated
over
time
and
it's
freshness
or
liveness
doesn't
matter
as
much
as
the
extended
visibility
so
like.
C
If
you
can
look
long,
if
you
can
look
over
a
long
time,
Horizon
you're
going
to
see
much
more
useful
information
about
what
the
workflow
looks
like
yeah,
yeah,
Trends
and
but
but
then
what
you
need
to
do
is
you
need
to
basically
make
decisions
about.
C
You
know
how
to
change
the
mix
and
the
activities
in
the
workflow
as
a
result
of
what
you
see
and
not
necessarily
you
know,
trigger
anything
automated
at
least
not
until
you
know
into
the
foreseeable
future.
Right
I
mean
the
things
that
we
need
to
do
are
fairly
simplistic:
you're,
gonna
and
they're
they're
fairly
standard
right,
I
mean
you're,
always
trying
to
lower
work
in
progress,
you're,
going
to
always
try
to
break
apart
the
work
to
simplify
it,
eliminate
dependencies
change.
C
C
So
it
really
is
not
a
real
time
need
it's
very
much
like
aggregating
data
and
visualizing
it
effectively.
So
it's
the
primary
challenges.
Access.
C
B
I
need
to
go
actually,
but
it's
a
very
interesting
discussions
and
I
start
to
understand
more
and
more
about
what
VSM
is
and
the
the
boundaries.
So
the
connections
between
cicd
domain
and
nvsm
to
continue
that
and
just
to
clarify,
there's
more
for
ourselves
and
for
for
both
of
us.
Well.
D
B
A
B
Appreciate
the
time
the
first
session
we
had
with
you
in
Helen
and
those
guys
some
some
months
ago
was
it
I
think
our
technology
was
quite
different,
but
but
we
are,
we
are
getting
there
I
think.
So
it's
and
I
think
that's
one
core
core
part
yeah.
C
It's
very
it's
very
rewarding
to
to
like
put
the
ideas,
throw
the
ideas
back
and
forth
and
throw
the
terminology
back
and
forth,
and
the
analogies
and
everything
else
yeah
so.
B
I'm,
looking
forward
to
the
next
time,
I
will
look
into
this
document
a
bit
more,
that
you've
shared
and
also
maybe
look
into
a
bit
more.
The
shorter,
we'll
see
if
I
find
time
for
that
where
we
are
getting
with
our
interoperability
initiative
there.