►
Description
For more Continuous Delivery Foundation content, check out our blog: https://cd.foundation/blog/
A
B
C
A
A
A
Yeah,
but
that's
the
one
that
points
to
the
event
vocabulary,
which
is
weird.
B
A
Good
stuff,
so
quick
update
on
on
my
side,
I've
sent
a
couple
of
pull
requests
today
for
andrea
and
to
review,
which
are
basically
like
refinements
in
the
vocabulary
removing
the
buckets
from
the
event
types.
I
remember
that
there
was
an.
C
A
I've
started
seriously
thinking
about
like
starting
implementing
something
because
we
need
some
kind
of
a
poc
at
some
point,
so
yeah,
I'm
just
trying
to
define
how
that
at
least
the
first
step
towards
like
a
plc
would
look
like
and
yeah.
I
was
planning
to
share
a
bit
of
that
during
this
meeting.
If
everyone
is
okay
with
that,
I
think
that
andrea
might
be
joining
later
today.
A
So
maybe
we
should
wait
for
for
him
to
join
and
see
where
he's
up,
and
he
has
if
he
has
any
news,
any
news
from
from
you
guys,
like
matthias
and
emil.
B
A
That's
that's
all
right.
I
mean
I'm
also
kind
of
like
trying
to
come
back
and
catch
up,
and
I
hope
that
I
got
that
like
sense
of
urgency
in
our
like
last
meeting,
like
the
team
meeting,
that
we
definitely
need
a
plc
so
yeah,
I
thought
that
yeah,
let's,
let's
start
creating
something
based
on
on
the
vocabulary-
draft,
which
is
pretty
simple,
but
at
least
it
can
be
used
for
kind
of,
like
the
basics
of
a
poc,
showing
interaction
between
different
things.
D
But
was
this
the
poc
is
this
kind
of
like
just
to
wrap
my
head
around.
It
was
this
like
a
converter,
for
example,
between
captain
or
tekton,
or
something.
A
A
By
setting
the
cloud
event
type
to
this,
you
know
to
this
type
and
and
then
of
course,
you
need
collect
the
code
that
will
create
that
structure
and
and
send
that
structure
to
another
system.
So
basically,
that's
where
I
I
wanted
to
start
by
just
creating
a
simple
library
and
the
cloud
event
structures.
A
A
You
know
fields
that
needs
to
be
filled.
For
that
specific
event.
As
soon
as
we
have
that,
then
we
can
use
that
library
to
say.
Okay,
tecton
is
emitting
these
events
and
captain,
for
example,
is
waiting
for
these
events,
so
we
can
translate
between
one
and
the
other
right
and
as
soon
as
we
have
that,
then
we
can
say
to
tecton8,
hey
techton.
If
you
emit
this
event
using
this
library,
then
captain
will
be
able
to
understand
you.
A
If
captain
is
you
know
receiving
those
events
by
using
the
library
or
using
any
other
implementation
of
that
library.
Basically,
by
reading
the
cloud
events
content
that
it's
been
sent,
I
got
kind
of
like
that
impression
from
alloys
from
from
captain
as
well
that
they
are
interested
in.
A
You
know
in
kind
of
like
seeing
that
cloud
event
format
and
having
a
say
on
the
format
itself,
so
by
having
a
small
library
and
kind
of
like
the
basic
data
for
each
of
these
cloud
events,
we
will
get,
you
know
them
involved
and
saying
yeah,
but
you're
missing
this
field
or
you're
missing
this
property
and
the
same
with
the
tecton
guys
and
with
the
same
with
every
other
project,
I
would
assume
you
know.
A
I
know
that
we
have
google
summer
of
code
now
happening
and
we
will
have
someone
creating
there
like
some
jenkins
events,
so
we
can
also
get
them,
including
involved
into
the
process
of
refining
these
events.
At
least
you
know
playing
around.
If
we
have
like
10
or
or
20
events
to
find
there,
then
we
are
talking
about
something
that
is
concrete
and
there
is
a
library
already
kind
of
like
an
sdk
allowing
you
to
create
these
events
and
programmatically,
send
them
to
different
tools.
A
Yeah
yeah,
it's
pretty
simple,
it's
kind
of
like
the
first
step,
I
would
say
the
next
and
you
know
again,
we
will
have.
We
will
have
the
chance
to
improve
the
vocabulary.
Documents
plus
have
the
cloud
mapping.
Events
mapping
in
there,
plus
a
small
library
that
we
can
just
share
with
people
and
people
can
just
submit
pull
requests
if
they
want
to
add
a
new
event
or
support
a
new
event
that
is
defined
in
the
vocabulary.
A
Once
we
have
that,
then
we
just
can
like
build
that
other
component,
more
like
as
a
demo
right
like
us,
just
showing
for
conferences
and
stuff
of
a
translator
between
different
tools
and
again,
the
tools
can
be
whatever
it
doesn't,
really
need
to
be.
Tucked
on
and
doesn't
need
to
be
kept,
and
just
I
thought
about
like
those
two
projects,
because
we
have
people
involved
from
those
two
projects,
but
we
can
have
any
other
project
like
jenkins
or
any
other
cic
pipeline
or
any
other
tool
that
it's
out
there,
that
we
can
demonstrate.
D
Yeah,
but
we
do
you
think
that
captain
would
listen
natively
on
the
on
this
new
event
type
or
should
be
converted
to
their
format.
A
I
think
that
nowadays
we
will
just
need
to
compare
to
tecton
format,
captain
format
and
any
other
tools
format,
because
of
because
of
time
restrictions
right.
I
know
that
captain
they
want
to
adopt
whatever
events
we
define,
but
they
have
a
tool
already
that
it's
already
listening
for
certain
event
types
and
they
because
they
have
like
a
real
implementation.
They
know
exactly
what
what
is
the
information
that
they
need
in
order
to
work
and
correlate
events
and
all
that
stuff?
A
D
Yeah,
so
basically
it's
it's
one
event
center
one
event
listener.
A
Yep
yeah
exactly
and
imagine
that,
like
that
plc
being
kind
of
like
the
router
between
these
components,
just
doing
the
translation,
which
I
think
that
captain
is
doing
already
internally,
I'm
pretty
sure
that
they
are
translating
a
bunch
of
stuff
for
like
their
integrations
right.
If
they
want
to
integrate
with
any
tool
now,
they
cannot
rely
on
their
or
you
know,
on
their
own
even
formats.
They
probably
are
just
picking
up
external
events
and
then
translating,
and
we
have
eric
hi
eric.
A
No
worries
no
worries.
We
were
talking
about
like
the
first
step
of
creating
a
poc,
and
I
think
that's
why
we
pinged
you,
because
you
mentioned
at
some
point
that
you
had
some
free
time
and
I'm
trying
to
define
that
and
trying
to
put
some
time
on
on
creating
kind
of
like
the
usual
steps
there.
E
A
A
Yep
cool
cool,
so
what
I
can
do
is,
while
we
talk
about
this,
because
I
know
that
we
have
kind
of
silicon
coming.
So
what
I
can
do
is
doing
this
week
or
early
next
week.
I
can
share
a
repository
with
just
the
basic
library,
defining
the
basic
cloud
event
types
and
maybe
eric
what
you
can
help
me
with
is
to.
E
Yeah
I
I
tried
reading
up
earlier
or
last
week.
I
guess
on
on
how
to
actually
define
a
schema
for
this.
So
my
take
was
that
somehow
the
the
schema
it's
sort
of
broken
into
several
parts
like
you,
have
a
schema
for
each
event
data
and
then
you
have
a
separate
schema
for
the
events
themselves.
I
guess
I
didn't
quite
understand
it.
I
guess
what
I'm
saying
is
I
I
would
need
to
read
or
find
a
proper
cloud.
E
A
Yes,
I
so
basically
what
this
is
a
tricky
thing
right
like
cloud
events,
doesn't
find
much
in
general,
so
you
can
like
extend
cloud
events
and
you
want
to
define.
I
think
that
I'm
not
entirely
sure
if
we
should
define
kind
of
like
a
schema,
I
mean
we
can
define
a
schema
for
the
extension
in
this
case,
but
then
we
can
just
use
you
know
json
schemas
or
whatever.
E
So
the
cloud
events
gamer
will,
for
instance,
define
that
there
should
be
an
events
type,
but
it
doesn't
really
help
us
define
what
event
type
matches
to
what
data
type.
In
some
sense,
the
event
itself
will
contain
that
information,
but
I
assume
at
some
point
there
should
be
a
schema
that
actually
says
what
events
we
have
exactly,
what
the
data
schema
is
used
for
each
event,
etc,
etc.
So
that
is
the
format
that
I'm
wondering
how
that
is
built.
E
A
Definitely
we
can
research
about
like
how
other
people
is
creating
their
own
cloud
events
if
they
are
right
like,
for
example,
I
would
expect
like
captain
to
have
that.
I
think
that
they
shared
kind
of
like
the
documentation
on
their
events
and
the
the
data
that
is,
is
expected
for
each
of
these
events,
and
we
can
take
a
look
at
what
tools
are
they
using
to
define
that
if
they
are
using
any
tools.
C
Yeah,
that
sounds
good.
So
for
what
we
do
for
tech
furnace,
I
mean
for
the
event
type
we
just
document
what
the
events
type
available
are
and
well
for
for
the
body,
it's
pretty
straightforward
in
the
sense
that,
because
tecton
is
based
on
crds,
so
customer
resources
every
time
a
custom
resource
triggers
an
event.
C
C
A
Sounds
like
yeah
sounds
like
a
kubernetes
way
of
doing
those
cloud
events
and,
I
think,
that's
kind
of
like
okay.
The
problem
is
that
in
this
case,
the
events
are
completely
separated
from
the
from
the
components
that
are
generating
them.
So
we
cannot
trust
that
there
is
a
kubernetes
resource
for
that.
A
So
andrea,
thank
you
for
joining
and
what
we
were
discussing
is
kind
of,
like
the
initial
steps
for
a
plc
and-
and
basically
what
I
was
trying
to
say
is
that
I
will
spend
some
time
in
creating
kind
of
like
that
initial
cloud.
Events,
mapping
with
the
stuff
that
we
have
in
the
vocabulary,
plus
a
small
library
in
gold
to
actually
create
those
cloud.
Events
right
like
it's
going
to
be
pretty
simple,
but
I
think
that
by
just
having
those
two
things
we
can
invite
people
to
review
the
event.
A
A
C
A
A
No,
I
think
that
there
is
no
problem
as
soon
as
I
can
just
send,
pull
requests
and
and
people
can
review
them
right
or
I
should
just
push
to
to
a
branch
or
something
like
that.
C
A
Yep,
whatever
works
better
right,
like
a
new
repo,
will
work
directory
directory
there
in
the
ripple
as
well.
So
I
can
just
submit
something
there
that
we're
going
to
start
reviewing.
I
would
appreciate,
I
know
that
eric
will
have
some
time
eric
if
you
want
to
just
jump
into
that
code
and
that
stuff
as
well.
That
will
be
great.
I
just
need
probably
a
few
days,
probably
until
monday,
to
clean
up
and
just
push
good.
Then
I
have
a
few
days
to
learn.
Go
because.
E
A
E
A
C
Sounds
good
that'd
be
great,
so
as
soon
as
we
have
that
I
could
also
try
and
see.
If
there
is
anything
I
can
do
on
tactile
side
to
emit
those
events.
C
C
Yes,
yes,
so
I
was
thinking.
Maybe
I
can
create
a
simple
controller
based
off
the
text
and
controller
which
the
only
thing
it
does.
It
sends
the
events,
but
I
can
physically
use
the
base
code
structure
for
the
main
controller
and
instead
of
sending
blank
cloud
events,
I
can
send
the
cloud
events
from
your
library.
The
only
the
only
beat
that
is
missing
is
kind
of
defining
some
kind
of
not
standard,
but
sometimes
yes,
I'm
some
kind
of
standard.
D
C
A
Make
sense
yeah
make
sense.
I
will
try
to
reach
out
to
captain
guys
to
see
kind
of
what
kind
of
integration
we
can
create
like
easily
right.
Like
the
same
thing,
if
we
submit
some
of
these
events,
where
do
we
put
kind
of
like
the
trans
transformation
layer,
so
they
can
just
get
these
events
and
do
something,
and
even
if
we
it's
like
a
small
interaction,
I
think
that
we
will
have
enough
to
show
that
this
is
valuable.
C
A
C
Yes,
I
started
looking
at
those.
I
think
they
they
look
okay,
so
if
yeah,
if
everyone
is
okay
with
with
those
with
me
going
ahead,
I
think
I
will
just
approve
them
and.
E
Yeah
I've
already
approved
them.
I
think
they
look
good,
okay,
good
okay
for
the
for
their
proof
of
concept.
Just
from
my
understanding
what
will
carry
the
the
events
between
tecton
and
captain?
That's
the
beauty
of.
A
E
I
was
thinking
if
we
use
some
sort
of
event
bots
like
like
nuts
or
or
something
like
that,
then
we
could
for
for
like
demoing
purposes.
We
could
also
attach
something
else
that
just
visualizes
the
events
that
are
sent,
so
you
can
see
that
okay,
techno
does
something
and
then
captain
does
something
else.
But
you
can
also
like
monitor
the
events
itself
on
some
sort
of
simple
dashboard
or
whatever
we
have.
A
That
sounds
good
yeah.
What
I've
used
before
is
basically
I
I
created
kind
of
like
a
k
native
broker,
which
is
basically
a
cloud
event
broker,
and
then
you
can
tap
into
into
that
broker
to
show
the
messages
that
are
flowing
right,
but
yeah.
We
can
definitely
use
any
other
tool
that
allows
us
to
monitor
cloud
events
that
are
being
sent.
E
A
E
C
Yeah
abstractions,
so
you
can
find
channels
you
can
have
in-memory
channel
or
you
can
use
it
on
top
of
kafka.
You
can
have
it.
That
could
be
one
one
way.
Yeah.
D
Tools,
I
shared
a
link
to
the
golden
sdk
for
cloud
events,
and
there
says
nuts
was
that
what
you're
talking
about
eric.
E
A
Really
good
yeah!
That's
why
I
can
like.
If
we
use
the
golan
sdk
for
cloud
events,
we
would
just
have
access
to
all
these
mappings
and
again,
if
you
use
k
native,
we
then
can
just
choose
any
of
those
things
like.
I
think
that
they
have
support
for
kafka
nuts
and
an
http
broker,
so
that
that
should
be
all
right.
But
again,
the
problem
here
is
the
complexity
of
the
demo
right
like
setting
it
up
and
showing
something
meaningful.
So
I
will
try
to
focus
on
that.
A
If
we
can
do
something
like
that
for
captain,
that
will
be
awesome
but
yeah.
I
will
that's
why.
I
think
that
if
we
have
something,
then
we
can
just
call
this
guys
and
say:
hey:
can
you
implement
something
quickly
to
understand
these
events,
so
we
can
just
demo
it,
but
first
first
we
need
to
do
something.
We
need
to
have
something
for
them
to
see.
A
E
No,
I
also
haven't
been
super
present
lately,
but
I'm
not
sure
if
andre
our
email
have
some
some
updates
from
the
the
sort
of
regular
sig
events
stuff
that
affects
us.
A
E
E
I
think
one
thing
I'm
not
saying
I
don't
think
it's
urgent,
I
think
much
of
it
can
wait
until
after
the
poc
is
done,
but
there
are
an
increasing
number
of
discussions
in
guitar
that
are
sort
of
not
getting
resolved
so
they're
just
getting
old,
which
means
we
either
had
to
have
the
discussions
again
at
some
point
or
or
we
I
don't
know
what
to
do
with
them.
So
maybe
maybe
that's
a
lesson
learned.
E
That's
rather
than
creating
too
many
discussions
at
the
same
time,
maybe
we
should
just
like
park
some
issues
and
then
create
a
discussion
for
it
later.
I
don't
know
how
to
handle
it,
but
right
now
there
are
probably
around
20
discussions
and
I
don't
think
more
than
one
or
two
of
them
are
sort
of
resolved
in
any
way.
A
Yeah
yeah,
that's
a
good
point.
We
we
are
good,
creating
discussions,
but
not
resolving
them
right.
Maybe
maybe
we
can
just
try
to
prioritize
the
active
ones
like
the
ones
that
were
created
in
some
way.
We
just
rank
them
and
then
just
discuss
them
in
like
in
the
regular
meetings
and
the
ones
that
are
related
to
the
vocabulary.
We
just
discussed
them
here,
but
the
ones
that
are
not
related
strictly
to
the
caller.
We
just
moved
to
the
main
meeting.
E
Yeah,
I
think
andrea.
I
mentioned
that
as
something
that
I
could
do,
which
I
haven't
done
is
to
like
really
try
to
figure
out.
What
is
the
minimal
set
of
questions?
We
need
to
answer
to
get
something
out
the
door
and
I
think
doing
the
poc
will
definitely
help
us,
identify
those
questions
and
focus
down
on
it.
So
I
think
it's
a
great
great
way
of
doing
it.
A
I
will
I
will
try
to
find
that
out
by
installing
it.
After
after
pushing
the
the
initial
you
know
commit
library,
I
will
try
to
find
that
out
because
yeah
I
haven't
used
it
and
I'm
guessing
that
it's
lightweight,
but
I'm
not
entirely
sure.
So
I
just
need
to
do
that
exercise
and
yeah,
but
if
not
I
mean,
I
think
that
if
not,
we
can
just
create
like
a
cluster
somewhere
and
just
put
it
there
not
sure.
A
D
I
haven't
used
it,
but
when
I
read
the
documentation
on
on
github
it
says
it
uses
this
k3s,
so
it
should
be
quite.
C
Cool
nice
yeah
techno
itself
is
also
quite
lightweight.
So
if
we,
if
we
have
a
lightweight
option
for
like
a
message
broker
or
like
a
lot
of
event
broker,
then
I
think
it
should
be
possible
to
run
everything
locally
and
have
a
nice
demo.
A
A
Good
stuff,
all
right,
let
me
see.
C
Another
thing
we
could
do
is
to
yeah,
I
don't
know,
maybe
it's
just
tilly
but
you're,
something
like
me
or
something
to
you
know.
A
You
mean
to
okay,
just
to
run
stuff
like
distributed
like
in
this
in
different
clusters
that
are
not
running
in
the
same.
C
Place
yeah,
it's
just
sorry,
just
brainstorming,
yeah
yeah.
A
A
A
Cool,
I
can
see
that
viva,
you
put
something
there
in
the
in
the
comments
that
you
are
going
to
share
plc
as
well.
I'm
wondering
what
that
poc
is.
If
you
can
share
some
more
details,
maybe
if
you
want
to
do
it
in
slack,
if
you
don't
want
to
do
it
here,
just
do
it
in
slack,
so
it's
not
like
deleted
when
we
finish
this
conversation.
A
A
E
A
Yeah
yeah,
it
should
be
not
that
large
cool
guys
any
other
topic
for
today's
meeting.
Regarding
the
vocabulary
again,
I
think
that
it's
all
about,
like
looking
at
the
events
that
we
have
defined
there
like
in
the
directory
and
start
thinking
about
the
specific
properties
that
we
want
to
add
and
the
kind
of
the
required
data
that
we
need
to
send
for
all
the
events
and
yeah.
As
soon
as
I
share
the
core
the
code,
I
will
expect
that
to
be
wrong
completely
wrong,
so
feel
free
to
send
comments
and
changes.
A
Maybe
something
that
can
help
me
andrea
is.
Can
you
share?
I
remember
that
you
share
at
some
point
kind
of
like
the
tecton
events.
Can
you
share
kinda
like
the
code
where
those
events
are
being
created
or
something
like
that
sure
yeah
yeah?
If
you
share
that
for
me
just
to
take
a
quick
look
and
make
sure
that
I
follow
some
of
the
same
patterns?
A
Is
there
any
other
topic
that
anyone
wants
to
share
or
or
talk
about
today?
Do
you
have
any
discussions?
We
can
go
through
any
discussions
regarding
the
vocabulary,
I
would
say,
because
if
not,
I
think
that
maybe
that's
kind
of
like
the
exercise
that
we
need
to
do
mathias
is
like
try
to
rank
or
try
to
in
some
way
prioritize
those
discussions.
D
On
a
previous
one,
we
had,
as
you
seen
in
the
protocol
there,
we
used
this
meeting
to
walk
through
a
couple
of
them
to
discuss,
so
that
was
kind
of
like
if
there
was
something
around.
If
it
was
a
discussion
that
was,
could
help
us
to
get
forward
with
the
puck.
A
D
I
was
more
a
question
if
you
had
something.
If
somebody
had
a
discussion,
I
thought
it
was
good
to.
E
I'm
not
sure
what
the
scope
is,
but,
but
I
think
when
it
comes
to
like
things
that
are
sort
of
like
a
lot
of
discussions
are
things
that
we
may
want
to
add
in
the
future
and
those
we
can
skip
for
now,
but
the
the
change
discussion,
which
I
think
you
matthias
brought
up.
Maybe
that
is
something
that
we
should
try
to
solve
right
now.
A
Yeah,
so
I
guess
that
the
the
definition
of
what
the
change
is,
if
we
yeah
yeah
yeah.
I
remember
that
discussion
and
I
would
definitely
go
and
try
to
add
some
comments
there,
because
I
do
see
this
being
kind
of
like
linked
to
the
poc
in
some
way.
Right.
Yes,
again,
if
you
have
a
pull
request,
triggering
a
pipeline,
that
would
make
sense.
E
Yeah-
and
I
think
I'm
not
sure
if
we've
allowed
time
for
it,
but
if
we
can
include
that
in
pc,
it's
a
very
powerful
way
to
like
show
just
any
regular
developers
that
okay,
I
can
actually
with
my
code
change.
I
can
make
a
lot
of
interesting
stuff
happens,
so
that
might
be
cool
from
that
perspective.
A
E
I
think
you
might
just
call
it
a
proposal
which
I
like
so
something
that
could
go
in
into
a
repository,
and
then
we
have
the
change,
which
is
when
something
actually
changes
on
like
the
main
branch
in
the
repository
and
then
we
also
have
a
commit
which
well
an
individual
commit
and
those
three.
It's
a
bit
tricky
to
draw
the
line
between
those
three
and
it's
important.
E
I
think,
for
the
vocabulary
that
we
do
have
clear
lines,
because
a
proposal
is
not
the
same
thing
as
a
commit
and
a
change
is
not
the
same
thing
as
a
commit
either
and
the
difference
between
a
proposal
and
a
change
it.
It
could
be
that
they
are
the
same
thing,
but
we
should
at
least
make
sure
that
in
the
vocabulary
we
we
know
what
we
mean.
Then
we
have
branch
as
well,
which
is
also
tangentially
related
to
at
least
a
change
in
in
some
systems.
E
A
Yeah
yeah
totally
but
eric.
I
don't
think
that
I
think
that
if
you
and
matthias
and
andrea
are
discussing
this
already,
I
think
that
sending
a
pull
request
to
the
right
file
in
the
vocabulary
will
help
to
move
the
discussion
forward
in
a
sense
that
we
can
have
a
proposal
for
what
do.
We
think
that
exactly
is
right,
like
change
proposal
and
all
these
new
terms
and
the
events
related
to
that,
and
then
we
can
just
make
a
decision
on
that,
because
I
guess
that
that's
how
kind
of
like
the
poc
will
move
forward
right?
A
E
Just
so,
we
don't
end
up
in
a
circular
way
of
working,
because
last
time
I
tried
creating
a
pull
request.
I
was
asked
to
give
everything
to
discussions
instead,
so,
but
I
think
what
I've
learned
from
that
experience
was
don't
create
big
pull
requests.
I
will
create
smaller
pull
requests
based
on
the
discussion,
so
I
will
happily
give
that
a
shot.
I
will
take
matthias
proposal
with
the
proposal
and
and
try
to
create
a
buddha
request
from
that.
A
C
Yeah
the
other
thing
about
changes
that
just
came
to
mind
is
I
mean,
if
you're
going
to
have
like
gitlab
or
github
hosting
the
the
code
that
you,
where
the
change
is
made.
They
have
custom
events,
their
own
type
of
events.
They
will
send
so
either
we
we
need
to
put
a
translation
layer
there
again
or
the
alternative
could
be
for
there
just
to
to
use
the
native
events
just
to
trigger
tecton
and
then
use
our
own
specific
cloud.
Events
between
tactile
and
captain
only.
C
I
guess
the
advantage,
if
had
our
own
events
and
translation
layer
on
the
change
as
well.
Then,
if
you
had
something
like
you're,
mentioning
eric
having
some
tools
that
receives
all
the
events
and
visualizes
what's
going
on,
then
it
would
get
also
that
the
change
event
so
in
that
sense
yeah.
That
would
be
nice,
but
we
need
some
kind
of
translation
later
than
that.
E
E
Then
there
are
tools
that
handle
changes
and
reviews,
like
probably
you
have
the
big
ones
like
github
and
gitlab
and
bitbucket
and
garrett,
maybe
maybe
there's
not
an
infinite
or
there
wouldn't
be
an
infinite
in
the
other
direction
either,
but
not
so
many
different
things
to
support.
So
maybe
it
could
be
okay
in
that
that
direction
to
have
some
sort
of
translation
layer
and-
and
we
can
focus
in
the
poc
and
the
other
parts,
but
convincing
garrett
via
plugin
to
send
our
cloud
events
should
not
be
super
difficult.
I
know
you
materials
have
something
similar.
D
Yeah
we
have
for
garrett
nicholl.
E
D
So
yeah,
if
you
have.
E
D
No,
I
was
actually
trying
to
switch
topic.
You
had
a
discussion
on
what
do
you
want
to
present
at
cdcon?
I
wondered:
are
you
satisfied
with
that
discussion
from
what
we
have
here,
or
it
should
be?
You
want
some
more
in
that
one.
E
So
the
the
actual
sort
of
time
slot
we
have
for
con
is
is
mostly
going
to
be
some.
It's
called
birds
of
a
feather.
I
think
some
sort
of
like
discussion
between
people
who,
who
have
some
knowledge
or
people
are
interested,
but
it
would
be
cool
to
have
sort
of
some
proof
of
concept
type
thing
to
launch
when
we
are
doing
that.
Talk,
oh
by
the
way
you
can
can
go
here
and
read
this
information,
and
here
you
can
see
things
actually
working
together.
E
D
Okay,
so
basically
we're
we're
kind
of
like
I
don't
know
how
we
want
to
deal
with
this.
Should
we
end
the
discussion
then,
or
should
we
add
in
information
and
close
it?
I
don't
know
how
these
discussions
work.
E
E
So
the
other
thing
I
wanted
to
bring
up
was
this
heavyweight
versus
lightweight
messages.
I
think
we
seem
to
be
in
agreement
most
of
us
that
lightweight
messages
with
links
is
the
way
to
go.
So
maybe
maybe
there
isn't
much
to
discuss,
and
we
can
just
say
that,
that's
what
we'll
go
for,
rather
than
trying
to
dump
absolutely
every
piece
of
information
we
have,
we,
we
will
send
links
to
where
people
can
fetch
more
information.
E
That
does
make
things
a
little
bit
tricky,
of
course,
because
the
the
information
that
is
available
in
the
link
to
more
information
will
probably
have
different
structures
depending
on
what
tool
you're
using
so
so
maybe
there's
something
to
to
think
about
there,
but
again,
maybe
not
for
the
poc.
A
Yeah,
I
would
definitely
leave
that
for
after
the
plc,
but
definitely
a
discussion
that
it's
worth
having,
I
think
yeah
definitely,
and
I
really
like
the
linking
presentation.
So
we
need
to
start
thinking
about
that
and,
as
emil
mentioned
at
the
very
very
beginning
of
the
conversation,
I
think
that
links
are
good
if
we
define
them
here
kind
of
like
the
vocabulary
as
like
extra
data.
E
I'm
not
sure
do
we
think
that
the
sig
interoperability
group
will
take
the
universe
in
the
direction
where,
whatever
those
links
are
pointing
to
is
more
likely
to
be
similar.
It
seems
to
me
that,
even
though
it's
a
massive
scope,
of
course,
but
at
least
that
seems
to
be
in
line
with
with
what
that
group
is
trying
to
achieve,
I
think
some
people
here
are
on
part
of
that
group
as
well.
E
So
basically
saying
that
if,
since
both
github
and
gitlab
and
garrett
has
commits
because
they
are
wrapping
git
repositories,
if
I
ask
those
three
systems
for
information
about
the
commit
like
how
different
would
that
be?
Is
that
something
that
this
interoperability
group
would
like
provide
guidelines
for,
or
recommendations
for,
or
something
like
that.
D
They're
at
least
discussing
it,
yeah
metadata
discussion,
and
I
there
I
guess
we're
we're
starting
to
look
in
or
I
have
I'm
not
gonna
be
part
of,
especially
that
investigation.
But
I
was
looking
to
xpdx
spdx
to
see
what
they
have
today
and
if
we
can
align
something
there,
because
there
is
a
protocol
or
a
actually
an
iso
standard
for
defining
package
information
he
started
out
in
in
license
handling.
D
I
think-
and
I
was
transitioning
in
and
they're
in
the
version
3.0
they're
looking
into
going
into
like
see,
I
see
the
domain
and
actually
getting
more
information
out
there,
and
so
that's
why
the
interoperability
is
going
to
be
looking
into
that
one.
So
hopefully
we
can
get
that
discussion
can
help
us
in
in
how
information
is
going
to
be
looking.
For
example,
if
you
talk
about
a
git
commit,
how
does
that
look,
and
so
on.
C
Right,
but
I
think
one
of
the
benefits
of
our
product
will
be
also
to
define
at
least
like
a
base
set
of
fields
that
should
be
in
the
events,
and
there
would
be
the
responsibility
of
the
translation
layer
or
of
the
search
system
directly
to
like
put
the
right
information
in
the
right
place.
C
I
mean,
if
you
want,
one
wants
to
dig
in
and
there's
more
information.
I
think
definitely.
This
would
be
helpful.
So,
in
my
mind,
when
I
was
thinking
about
light
weight
versus
heavyweight
messages,
it
was
more
like
yeah
thinking
about
actually
like
more
heavyweight
kind
of
artifacts
like
log
files
or
junit
xml.
C
This
kind
of
artifacts
and
and
those
would
be
like
links
to
text
files
binaries
that
you
can
download
and
the
format
could
be
standard.
If,
like
you,
have
a
jxml
standard
that
is
defined
in
that.
A
D
When
it
comes
to
sites
and
so
on
cloud
events
is
actually
recommending
users
to
keep
the
events
compact,
and
so
it
recommends
not
to
embed
large
data,
but
it
wants
to.
It
recommends
to
link
it.
If
you
read
enough,
when
I
send
the
link
to
its
size
limits
when
your
time
is
the
last
paragraph
paragraph
there
and
they
talk
about
that,
if
you,
if
you
link
to
information,
you
can
also
have
this
differentiation
of
access
control.
D
So
you
can,
you
can
publish
an
event,
but
you
don't
need
to
publish
the
log
because
to
get
the
log
you
need
to
have
a
login,
but
then
you
can
send
events
pretty
widely,
but
you
don't
need
to
send
out
information.
Detailed
information
about.
E
Log
yeah,
I
remember
reading
that
that's
true
yeah,
so
we
will
have
support
for
from
cloud
events
for
having
smaller
events
with
links.
That's
good.
D
Just
for
protocol
correction,
so
on
in
the
last
meeting
from
20th,
there
was
a
bunch
of
actions
and
how
do
we
want
to
handle
if
you
resolve
one
of
them.
A
Let
me
take
a
look
at
what
were
the
actions,
because
I
was
in
that
meeting,
so
I
not
entirely
sure
participants.
Oh
yes,
so
we
have
all
those
actions.
Okay,
I
I
think
that
the
actions
can
be
marked
now
right,
like
you,
can
just
mark
the
ones
that
were
done,
and
I
think
that
if
we
still
think
that
we
have
those
actions
to
be
executed,
we
should
move
them
to
the
new
meeting
right.
A
So,
basically,
what
I'm
doing
is
I'm
just
copy
the
sections
like
the
action
sections
to
the
to
the
to
today's
meeting,
to
the
notes,
and
if
you
resolve
the
action
you
can
just
mark
it
with
the
check,
because
I
think
that
they
like
they're,
using
checks
to
to
mark
that
and
and
then
the
ones
that
were
not
done.
We
should
move
them
to
today's
meeting,
so
there
are
actions
for
the
next
one.
A
E
Yeah,
I
just
posted
a
reply
to
to
the
what
we
want
to
present
that
cdcon
discussion.
I'm
not
sure
if
that
was
what
you're
looking
for,
but
there
I
have
what
I've
understood
at
least
perfect
yeah,
but
I
will
just
yeah
I
I
will
do
whatever
I
can
to
learn
a
bit
of
go,
but
then
I
will
wait
for
your
your
patch
or
or
commit
that
you
wanted
to
send
me,
and
I
will
look
into
that.
That
sounds
fun.
Yeah
great
great
awesome.
A
Okay,
I
need
to
jump
to
another
meeting
so
guys.
Thank
you
very
much
and
see
you
in
a
couple
of
weeks.
I
will
definitely
share
the
links
and
basically
ping
you
eric
as
well
in
the
in
the
channel
and
andre.
If
you
have
time
to
take
a
look
at
that
as
well.
That
would
be
great
because
again,
I'm
pretty
sure
that
it's
not
going
to
be
perfect.
So,
whatever
suggestion
on
style
or
whatever
else
that
you
see
that
we
can
just
quickly
adapt,
let
me
know
sure
great.