►
From YouTube: CDEvents Working Group (EMEA/APAC) - June 19, 2023
Description
For more Continuous Delivery Foundation content, check out our blog: https://cd.foundation/blog/
B
B
D
Yeah
yeah
I'm,
pretty
good
I,
was
over
actually
and
I
was
over
in
Merrimack
and
then
Fidelity
had
a
big
room
planning
session,
so
bitjet
lights
getting
back
on
the
Friday
morning,
but
good
session,
though
busy
week.
C
Okay,
shall
we
get
started
with
the
meeting.
B
Only
got
10
minutes
today.
Sorry
because
I've
got
to
go
to
another
meeting.
C
Okay,
then,
let's
get
started
so
welcome
everyone
to
City
events,
working
group,
June
19th
today
and
feel
free
to
sign
yourself
in
if
you'd
like
to
share
the
link
in
the
chat.
C
I
got
a
few
thanks
on
the
agenda,
not
many,
but
since
you
only
have
10
minutes
Brad
is
there
anything
in
particular
that
you
wanted
to
discuss?
First
before
we
go.
B
On
yeah,
just
to
update
on
the
GitHub
action,
the
facility
event,
yes,
and
possibly
just
throwing
an
idea
out
for
you
for
something
in
the
spec
that
I
believe
could
be
in
scope.
I'd
I'd
be
interested
on
your
thoughts,
which
I
can
share
as
well.
B
Yes,
I
can
yes.
B
Yeah
so
I
guess,
like
so
update
on
the
GitHub
action,
we've
made
a
mock
endpoint,
so
we're
now
ready
to
you
know
to
start
testing
it.
We've
done
the
code
and
now
we're
just
ready
to
start
testing
different
examples
and
and
and
get
get
it
ready
for
that
sort
of
first
version
and
then
the
next
meeting
we
can
present
that
back
and
then
four
four
City
events
I
was
going
to
wonder
about
the
so
if
I
get
the
spec
up
on
my
computer,
so
we're
wondering
at
work.
B
If
we
can
so
we
have
in
jarrah
we
we
track
our
features
right.
So
this
is
sort
of
mapping
the
life
cycle
from
idolation
right
to
delivery
so
right
to
the
end.
So
essentially,
what
we're
wondering
is:
is
there
any
scope
for
something
like
issue
created
or
feature
created,
or
some
sort
of
link
that
will
you
know
we
have
pipeline
run,
builds
Repository
environment
change.
Change
is
almost
close,
but
what
are
your
thoughts
on
adding
something
one
step
back
that
will
go
like
when
this
feature
is
come
from
an
idea
into
development.
G
I
would
say
this,
so
it
features,
I
will
probably
call
them
issues,
yeah,
I,
think
so
and
actually
I
mean
issues
could
be
both
features,
but
they
could
also
be
trouble
reports,
for
example,
connected
to
some
incident.
Yes,
it
would
guess
to
see
when
the
incident
has
been.
This
is
into
this.
The
CIO
City
pipeline
before
it's
actually
resolved.
B
Yeah
because
then
I
could
go.
You
know
this
issue
or
this
document
then
that
could
then
progress.
You
know
into
the
rest,
where
you
know
we
make
a
repository.
We
do
a
you
know
the
pipeline,
that
environment
created
all
those
things
and
then
we
could
track
right
from
the
start
right
to
delivery
and
then,
if
I,
if
I
tell
that
to
my
boss,
I
would
be
able
to
adopt
see
the
events
as
well,
because
we
do
very
well
in
jira
with
that.
G
Yeah
and
I
guess,
but
when
you,
when
you
push
a
change
to
to
your
Source
control
system,
you
also
often
not
annotate,
that
with
some
kind
of
jira
tag
or
something
in
many
cases,
I
believe
yeah.
A
B
Yeah,
because
when
we
we
use
brunt
and
we
use
our
jira
ticket
for
branch
and
PR
titles
in
in
like
the
semantic
version
than
the
ticket
but
yeah
I
guess,
I
was
also
playing
around
with
a
lot
of
open
Telemetry
as
well.
It's
almost
like
a
trace
ID
that
would
sort
of
do
the
propagation
right
through
the
life
cycle
of
that
feature
or
that
issue.
B
Easy
as
just
putting
the
first
ID
on
to
the
jira
ticket
and
having
a
custom
field,
there
would
be
fine
as
well,
but
I
can
think
about
it
more.
If
you
think
that
that
could
be
in
scope
and
then
maybe
maybe
start
a
thread
in
slack,
and
we
could
talk
about
it
or
maybe
I
can
do
what
Ollie
did
and
just
sort
of
make
a
really
loose
draft
proposal.
C
Yeah
so
I
think
I
agree
with
Emil
I
think
this.
This
isn't
scope
and
I
wanted
to
to
point
out
a
couple
of
things
that
have
been
going
on
that
are
also
relevant
to
this.
So
the
first
one
is.
We
started
a
collaboration
with
the
value
stream
management,
interoperability
group
from
the
Oasis
organization
and
they're,
basically
they're
interested
in
having
an
interoperable
way
to
define
yeah.
So
the
events
in
the
value
stream
management
area,
so
companies
like
atlassian
seriousness,.
A
C
Part
of
that
group,
so
their
scope
is
wider
than
what
CD
event
is
today,
and
we
are
not
necessarily
going
to
say
that
we
are
going
to
expand
the
City
events
to
do
whatever
they're
going
to
do.
But
this
certainly
is
a
good
amount
of
overlapped
and
we
already
discussed
in
the
process
of
information
issues
in
the
area
of
incidents,
so
I
think
certain
area
of
the
scope
could
be
included
in
City
events
for
sure
so
and
that's
something
I
think
that
is.
C
That
is
relevant
for
you
specifically
for
this
use
case,
because
you
were
mentioning
jira
and
issue
tracking
system.
So
that's.
If
we
can,
if,
like
that
group,
would
adopt
City
event
as
a
standard,
then
it
would
mean
that
CD
events
could
be
implemented
in
products
like
adolescent
products
and
similar
products
to
generate
those
events.
A
C
Other
work,
which
I
think
is
relevant
because
you
mentioned
like
connecting
the
events
or
tracing
for
good
events,
is
that
we've
been
discussing
about
how
to
connect
different
events
with
each
other
and
that's
a
work
that
we'll
be
discussing
some
for
some
time
and
then
from
Apple
is
preparing
a
proposal
that
hopefully
we'll
be
able
to
to
review
soon
and
that
we
want
to
include
at
least
in
the
first
iteration
in
the
next
release.
So
I
can
point
you
to
the
you
probably
find
some
links
in
the
meeting
notes,
but.
E
A
B
Okay,
cool
I'll
wait
for
the
proposal,
then,
and
I'm
sure
it
will
have
some
good
stuff
on
it.
Yeah.
C
But
by
all
means,
I
think
if
you
want
to
start
I
mean
like
describing
the
use
case
that
you
have
in
mind
the
idea,
Indonesia
and
start
the
discussion
and
it
in
parallel.
I
wouldn't
wait
on
other
proposals
for
doing
that.
I
think
it's
absolutely
valuable
to
do
that,
because
it's
it's
more
than
just
connecting
events
and
I
think
yeah.
It
also
something
that
we
can
bring
up.
Then,
when
we
have
Steve
from
the
bsmi
joining
the
group,
so
we
can
say:
okay.
B
All
right
yeah,
so
that
was
my
updates
I
had
overlapping
meetings.
I
had
to
run
out
of
that
one
just
to
quickly
give
you
an
update,
but
next
time
we'll
be
ready
to
present
the
GitHub
action
work
so
far,
and
then
we
can
get
alignment
and
then
keep
going
for
the
next
phase
of
it.
A
C
Okay,
I'm
trying
to
take
some
notes,
I
mean
maybe
you
can
may
want
to
complement
what
I
wrote
there
in
your
comments.
I,
don't
think
it's
quite
complete.
C
All
right
anything
else
or
any
other
comments
on
these
two
topics
from
from
Brad.
C
All
right
so
I
then
going
back
to
from
the
agenda
I
in
terms
of
action
items.
The
only
thing
I
had
noted
done
was
for
better
arrange,
a
meeting
with
SB
Network
team
or
demos.
The
work
Challenger
has
done,
I,
don't
know,
I,
don't
see
I'm
an
original
under
in
that
meeting,
so
I,
don't
think
we
we
have
any
update
on
this.
C
C
Project
which
is
a
artifact
repository,
it's
a
graduated
project
in
the
cncf,
so
I
presented
this
proposal,
which
is
not
fully
complete
yet,
but
it's
about
well
I
presented
a
proposal
and
then
I
think
it
was
received
quite
well
and
they
asked
me
to
make
a
formal
proposal,
which
is
this
PR,
that
I
created
now
and
yeah.
C
So
one
of
the
things
in
this
proposal
is
how
existing
events
from
Harbor
map
to
City
events,
so
the
harbor
is
a
number
of
different
events
that
are
available
and
I
think
only
push
artifact
as
we
have
on
City
Event
side.
D
C
I
believe
a
lot
of
artifact
repositories
include
scanning
capabilities,
so
this
probably
makes
sense
for
interoperability,
point
of
view,
I'm
less
sure
about
things
like
orders,
replication
tag
retention.
This
is
something
we
should
include
in
City
events
or
not,
but
yeah
I
didn't
want
necessarily
to
start.
The
discussion
here
just
wanted
to
provide
this
update,
but
if
anyone
has
votes
and
would
like
to
share
in
this,
please
let
me
know
I
included
here
a
link
to
the
issue
they
opened.
G
I
guess
it's
always
easy
to
to
dream
about
any
kind
of
event
like
if
it
sounds
from
another
system.
I
think
we
need
to
be
somehow
make
conscious
decisions
when
adding
a
lot
of
new
events
for
what
use
cases
they
would
be
interesting
for
and
if
they
are
really
CD
related.
So
let's
say.
G
One
obvious
idea
is
that
City
events
should
be
sent
when,
when
there
is
some
kind
of
use
case
saying
that
we
can
trigger
an
action
in
a
cicd
context
based
on
that
event,
but
we
also
know
that
it's
for
it's
for
observated
and
not
just
for
triggering,
of
course.
So
whatever
is
relevant
to
observe
from
a
sort
of
CR
series
systems,
let's
say
could
be
I
guess
for
syrians.
C
Yes,
indeed,
that
was
kind
of
my
a
similar
line
of
thought.
I'm
not
sure
quota
is
relevant
for
I
mean
it's
relevant
for
maybe
relevant
for
triggering
some
kind
of
administrative
or
notification
workflow,
but
not
the
cicd
workflow,
specifically
in
terms
of
observation
it.
It
might
provide
additional
information,
though,
about
why
a
certain
work
with
the
ICD
workflow
stopped
or
failed,
but
yeah
then
again,
I'm
not
sure,
and
we
need
to
draw
a
line
somewhere,
but.
F
C
Maybe
something
that
I
wanted
to
mention.
I
forgot
on
this
proposal.
Sorry
Harbor
proposal
in
terms
of
implementation,
I
think.
C
What
I
proposed
is
there
we
start
doing
an
initial
implementation
on
Harbor
Side,
that's
with
cdbank0.3,
and
that
means
only
the
push.
Artifact
event
is
there
and
then
once
we
have
more
implement
events,
respect
changes
required,
possibly
in
0.4,
then
we
can
iterate
and
do
like
larger
or
full
implementation
there
for
producing
events.
G
G
At
the
moment,
I
haven't
succeeded
in
finding
anyone
who
can
be
represented
yet
but
I
guess
we
have
similar
requirements
that
are
similar
needs
of
events
from
manufacturers
from
order.
C
Yes,
yes,
indeed,
so
I've
not
spoken
directly
from
with
the
this
product
owner
on
jfrog
side,
we
spoke
during
the
meeting
I.
Think
the
the
there
was
only
an
update
from
Laurie
that
she
had
some
more
internal
meetings
and
things
seems
to
be
promising
proceeding.
Well,
so
that's
that's
kind
of
the
the
only
update
we
discussed
about
this
in
the
last
meeting
with
with
Ben
and
the
page
then
reiterated
also
Apple's
interest
and
having
CD
events
supported
by
archifactory
and
offer
to
you
know,
join
the
conversation
as
well.
C
Yeah
I
think
even
that
easy
was
invited
to
the
initial
meeting
at
nailed
to
to
I.
Don't
know
if
you're
aware
of
this
I
think
Jared
was
invited
to
to
the
meeting.
So
we
have
some
near
the
meeting
with
jfrog
product
owner
discussing
to
discuss
and
present
the
value
that
architecture
kit
get
from
implementing
CD
events
and
because
both
Fidelity
and
apple
expressed
interest
and
I
believe
like
Emil,
was
saying
Ericson
as
well
uses
artifactory
and
they
have
a
support
for
I
feel
there.
C
G
I
think
it
would
be
powerful
to
have
a
simple
suggestion
for
what
Factory,
as
we
have
now
for
Harvard,
and
to
show
one
on
the
interoperability
perspective
of
City
events
as
well,
that
we
can
have
two
different
artifact
repositories.
Sending
the
same
events
and
consumer
could
could
see
those
events,
regardless
of
who
sends
them.
G
C
Yes,
I,
don't
know
that
we
have
a
commitment
yet
from
jfrog
that
they
would
like
to
do
with
implementation,
but
maybe
they
will
be
open
to
to
discuss
such
things
for
sure,
yeah,
okay,.
C
But
what
I
can
do
I
mean
I
can
to
start.
I
can
also
share
this
work
that
we're
doing
with
Harper
with
Laurie,
so
that
she's
aware
about
it
as
well
yeah-
and
maybe
she
can
relay
this
information
back-
that
this
conversation
is
relevant
for
them
so
and
that
they
they're
very
welcome
to
join
the
conversation,
Upstream
I
would
say,
and
so
we
can
discuss
what
what
events
make
most
sense
to
have
any
standard.
G
E
C
D
Exactly
what
kind
of
events
do
they
do
they
support
today
and
what's
what's
important
to
them,
I
suppose
and
I?
Don't
know.
If
that's
the
point,
that's
something!
Maybe
not
that's
more
question
for
are
more
in
interest
for
me
just
trying
to
understand
this
space
more
so
maybe
for
them,
but
I,
just
just
maybe
a
meal.
Just
take
you
just
to
breathe
forward.
You
know
my
background,
so
infidelity
we're
obviously
incorporating
CD
events.
C
D
You
know,
obviously
you
know
and
I
have
the
benefit
of
not
knowing
enough
about
this
to
ask
these
kind
of
questions,
but
it
seems
like,
like
a
no-brainer
that
you
would
you
would
want
to
use
just
for
interoperability.
You
would
want
to
use
this
spec
unless
you
had
some
bespoke
and
requirements
that
you
know
precluded
you
from
using
this.
So
that's
why
I'm
kind
of
wondering
about
these
questions
and
yeah.
C
Yeah,
so
in
the
in
the
discussion
with
with
jfrog,
we
had
I
think
product
manager,
the
owners
from
two
components
of
artifactory
one
is
the
Regis
one
was
registry
and
the
other
one.
There
is
a
pipeline
component
as
well,
so
I
think
the
the
current
the
event
from
the
CD
event
specification
that
we
have
today
that
would
be
relevant,
probably
for
them
or
the
artifactory.
Sorry,
the
artifact
published
event
that
we
have,
that
is
designed
specifically
for
registries
and
possibly
I.
C
C
But
yes,
I
think
it
would
be
very
interesting
for
artifactory
people
to
then
join
the
conversation
Upstream.
So
we
can
discuss
what
what
kind
of
events
they
they
have
today
if
any
and
what
it
would
make
sense
to
include
in
this
standard
in
addition
to
what
we
have
today,
so
we
can
more
the
more
voices
we
we
have
in
the
conversation
of
course,
the
the
better
decision
we
can
do
in
terms
of
internal
probability.
C
All
right
any
more
thoughts
on
this.
C
No
okay,
thanks
a
lot
for
the
input,
so
I'll
keep
you
posted
on
on
progress
for
that.
So
there
was
an
approval
on
the
on
the
proposal
already
I
think
there
is
some
more
detail,
design
that
I
need
to
do,
but
yeah
I'll
keep
I'll,
keep
you
posted
on
progress
there.
E
C
C
And
the
other
thing
if
we
have
time
or
if
you
prefer
to
just
look
at
that
first,
a
colleague
from
from
IBM
created
a
use
case
about
parent
child
pipelines
right
CD
events,
so
I
wanted
at
least
to
to
call
it
out.
So
if
people
want
to
to
have
a
look,
maybe
we
can
discuss
it
next
time
and
then
start
in
commenting
on
it.
C
Making
some
diagrams
I
think
the
the
use
case
is
to
help
the
kind
of
orchestration
between
parent
and
child
pipelines
for
events
using
City
events,
but
yeah
some.
Some
of
the
comments
there
seems
to
indicate
a
more
like
imperative
events
pattern
again,
so
I
want
to
try
and
describe
how
we
can
do
this
kind
of
use
case
for
declarative
events.
Instead,.
C
Yeah,
my
feeling
is
more
towards
the
second
but
I'm,
not
100
sure
so
I'm
still
going
through
it's
it's
quite.
A
C
So
I
think
there
is
a
very
describe
this
discussion
about
this.
But
my
feeling
is
more
about,
like
child
pipeline
orchestration
like
to
break
down
a
large
pipeline
into
pieces
connected
through
events,
as
opposed
to
necessarily
having
like
upstreamed
Downstream.
C
G
I
guess
that
was
triggered
on
the
dependency
management
term
there,
which
is
then
the
other
one
I
mentioned,
and
that's
related
to
an
issue.
I
wrote
some
time
ago,
a
year
ago,
issue
number
39
yeah.
A
C
A
The
provinces.
G
But
then,
if
it
is
really
apparent
child
relationships
within
a
popular
execution,
then
I
guess
we
we
talk
about
hierarchical
or,
like
pipeline,
run
events
connected
to
other
pipeline
relevant
Society.
It
would
be
a
hierarchical
accepting
of
510
events
that
events
rather
oh,
that's
depends
I'm,
not
sure
if
we
have
an
issue
for
that,
as.
A
C
Yeah,
so
we
we
have
a
concept
of
Pipeline
and
task
I
guess
you
could
encapsulate
them
within
each
other,
but
I
think
the
point
is
I
started
making
and
on
that
issue.
Is
that
since
the
what
what
we
do
in
CD
events
is
more
declarity
events
saying:
okay,
this
happens
and
not
requesting
something
else
to
happen.
So
you
don't
really
know
Upstream
who
is
going
to
react
Downstream
so
typically
that
decision
on
whether
to
react
and
what
to
do
with
the
reaction
is.
A
C
So
from
the
parent
pipeline
execution
point
of
view,
they
won't
really
know
about
the
downstream
execution,
so
they
cannot
really
wait
for
Downstream
pipelines
execution
to
complete
for
the
Upstream
pipeline
to
complete.
So
it
wouldn't
really
be
encapsulated.
C
It
would
be
more
like
collated
to
each
other,
so
one
triggering
and
arbitrary
number
of
other
pipelines,
potentially
and
I
think
then
we
could
I
was
planning
on
creating
some
diagram
showing
how
then
the
the
idea
of
like
collecting
all
the
events
in
one
place
and
defining
policies
could
be
used
and
to
trigger
the
next
step
in
the
workflow.
So.
D
C
So
you
could
collect
all
these
events
and
then
you
could
never
could
have
a
policy
to
say,
okay.
Well,
when
all
these
required
checks
are
executed,
then
you
can
trigger
The
Next
Step
on
the
next
pipeline,
which
is
a
deployment
pipeline.
For
instance,
we
could
be
even
managed
by
a
different
tool
that
from
the
tool
that
executed
the
build
pipeline,
okay
and
then,
if
you
add
another
check-
and
you
want
that
check
to
be
required,
then
you
would
need
to
alter
the
policy.
C
But
you
wouldn't
need
to
alter
the
source
pipeline
if
it
makes
sense
yeah
yeah,
which
is
one
of
the
issues
that
I
know
they're
struggling
with
on
IBM
side,
that
they
have
this
very
huge
pipelines
with
compliances
and
so
forth,
and
every
time
there
are
things
that
needs
to
be
changed.
So
everything
needs
to
go
back
into
the
big
Pipeline
and
that
needs
to
be
modified.
And
then
you
know
all
the
teams
that
use
that
large
Temple
pipeline
if
you
have
to
go
through
the
update
of
that
pipeline.
C
A
G
I
was
more
thinking,
I'm,
not
sure,
if
that's
the
same
thing,
but
if
you
would
consider
a
I
mean
when
you
start
implementing
your
CI
CA
club,
we
say
you
might
have
a
quite
simple
setup
with
just
a
few
steps
in
it
and
then
maybe
you
have
a
test
step
that
is
quite
rudimentary,
but
just
very
small
thing,
but
then,
as
time
goes
by,
it
might
increase
that
test
a
step.
G
The
test
activity
and
eventually
you
might
introduce
new
test
tools
and
such
and
maybe
those
that
test
type
itself
would
constitute
its
own
pipeline
full
pipeline
run
in
an
external
tool.
Maybe
that
you
just
don't
want
to
incorporate
within
the
top
pipelines
website.
So
then
it
would
be
like
you
pipe
down
the
Pipelines.
A
G
C
I
think
that
there
are
two
difficult
difficulties
with
doing
that
with
CD
events.
Today,
one
is
you
can.
How
can
you
trigger
the
child
by
pulling
from
The
Source
pipeline
for
City
events,
because
then
people?
Is
it
a
declarative
event
or
is
it
you?
You
want
to
have
a
van
that
says
start
this
pipeline.
C
We
don't
have
those
today,
so
you
would
say
you
will
need
to
say
something
like
the
artifact
is
packaged,
and
then
there
is
a
test
system
that
receives
that
events
and
knows
that
okay
yeah,
so
the
artifact
is
ready.
Now
so
I
will
trigger
the
tests.
C
Other
things,
then,
is
then,
from
parent
pipeline
point
of
view.
What
does
the
completion
of
the
pipeline
means
and
success
of
the
pipeline?
So
if
you
want
to
incorporate
the
result
of
the
testing
pipeline
inside
the
original
pipeline,
you
need,
as
you
say,
some
way
to
to
go
back
and
some
way
for
the
parent
pipeline
to
know
which
are
the
children
who
are
the
children,
and
you
know
have
like.
G
Yeah
yeah
for
sure,
but
I
mean
if,
if
the
pipeline
is
orchestrated
and
not
really
event
it's
trigger,
for
example,
you
have
a
what
do
they
have
a
test
step
in
your
pipeline,
executing
certain
test
activity
for
you
and
then
you
may
be
replaced
if
that's
a
or
a
basket
or
something
just
you're
running
some
tests.
But
then,
if
you
replace
that
with
a
full-fledged
test
system
which
has
its
own
pipeline
in
it,
but
you
call
it
directly-
it's
not
eventually,
but
you
call
it
from.
G
Pipeline
so
more
or
less
replace
your
tests,
your
home,
a
hard-coded
test
script,
with
your
call
to
some
external
tests
orchestrator,
which
has
its
own
pipeline
defined
for
it.
Maybe,
and
then,
when
you
when
that
tests
orchestrator
has
finished
its
testing
task,
it
will
just,
of
course,
return
control
to
the
decoder,
which
would
then
continue
the
pipeline
on
the
higher
level,
and
you
would
like
to
have
some
way
to
to
refer
the
events
sent
from
that
test
orchestrator
onto
the
top
level
pipeline
orchestrator.
If
you
okay,
what
I
mean.
C
Yeah
I
think
this
is
maybe
use
case,
but
it's
worth
and
calling
out
again
it's
yeah.
It
seems
possibly
related
also
to
to
that
issue
that
I,
142
and
I
think
it's
again
it's
relevant
for
for
the
connecting
events,
so
there
should
be
something
in
this
pack
that
says
well,
if
you're
triggered
with
a
certain
context,
you
should
include
that
context
in
the
event,
and
so
that
allows
you
to
the
link
things
back
right.
C
C
C
Where
we
have
two
page
about
two
pages
of
events,
sorry
of
issues,
can
we
start
from
the?
What
does
it
matter
really:
okay,
I
guess.
The
first
one
is
the
one
that
I
just
created.
Events
for
the
artifact
repositories.
C
G
I
guess
we
could
get
started
with
your
start
by
published
event
for
now,
just
to
make
sure
that
I
mean
the
infrastructure
is
there
and
the
event
can
be
sent
and
then,
of
course,
it
can
be
expanded
later
on.
Unless,
however,
thinks
that
it's
not
worth
much
to
have
that
event
without
having
some
more
events.
C
Yeah
no
I,
don't
think
so.
I
mean
at
least
in
The
Proposal
I
I
made
is
to
start
with
that
event
and
then
to
continue
from
there.
There
might
be
additions
that
we
can
do
even
to
the
artifact
boost
event
to
match
what
they
do
today
or
to
give
more
information.
So
that's,
maybe
also
part
of
the
discussion
or
maybe
I
could
create
a
separate
issue
for
those.
G
C
A
E
A
F
C
Well,
it
not
necessarily
started
that
I
guess
part
of
the
discussion,
so
we
could
not
it.
It
depends.
So
it
depends
on
the
test
system
that
you
use
and
I
mean
there
are
different
semantics.
So
in
some
cases
you
can
have
like
in
Python
you
can
have
a
discovery
step
that
discovers
the
test
and
you
can
have
skip
happening
at
Discovery
level.
C
You
can
have
skip
declared
in
the
beginning
of
a
test.
So
actually
the
test
is
discovered.
Then,
when
you
executes
based
on
the
context
that
is
available
at
the
execution
time,
it
decides
that
it
won't
run
because
it
is
running
on
I,
don't
know
the
wrong
operating
system
architecture.
Whatever
reason.
C
We
we
can
describe
I
think
as
part
of
this.
Maybe
it's
worth
refining
what
what
is
the
expectation
and
in
these
cases
and
whether
let's
keep
should
be
it
could
be
like
a
dedicated
predicate
or
it
could
be
so
that
if,
if
an
event
is
not
started
at
all,
you
could
send
say
event
was
test
was
kept
completely
or
we
could
say
started
and
then
finish,
but
it's
keep
raising
or
could
have
started
and
feed
and
skipped
and
not
ever
finished.
I
guess
it
depends
a
little
bit.
C
We
have
a
problem
in
the
CI
system,
where
certain
PR
would
suddenly
make
a
very
large
number
of
tests
skipped
and
then
the
test
job
would
run
successfully,
and
so
then
the
reviewer
could
not
realize
that
certain
tests
were
now
escaped,
not
the
author,
so
the
pr
is
merging.
C
We
certainly
lose
a
certain
amount
of
coverage,
and
you
you
can
capture
that,
of
course
for
coverage
reports
which
then
we
improved
upon,
but
also
if
we
had
like
historic
information
about
execution
of
tests,
you
could
see
that
a
certain
test
is
executed
and
then
is
skipped
and
that
is
different
from
it's
executing
and
then
it's
not
there
anymore,
because
it's
not
there
anymore
could
be
that
the
test
is
actually
been
removed.
C
But
yeah
so
I
think
that
there
could
be
value
in
in
navigate
I've
excluded,
but
what
I
suggested
then
we're
done
is
because
they
seem
to
have
use
case
for
this
and
maybe
make
a
proposal.
Then
we
can
continue
the
discussion
on
the
proposal.
E
E
C
I
guess
in
one
potential
output
is
that
we
decide
not
to
do
this
at
all,
but
at
least
we
we
do
discussion
about
it.
You
complete
the
discussion,
but.
A
A
F
E
A
C
G
G
D
The
knowledge
is
gents,
I
have
to
drop
further
meeting,
but
this
was
definitely
very
helpful
for
me
and
very
informative.
So
thank
you.
C
D
A
C
G
Think
it
does
so.
That's
that's
yeah
the
time
needed
I
guess
we
can
probably
have
some
timer
that's
this
afternoon
as
well
on
the
City
events
meeting.
If,
if
nothing
else
spring
comes
up,
I
haven't
thought,
unfortunately,
not
anything
yet
so,
okay
suffering
from
a
bit
of
a
headache,
so
I'm
not
sure
if
I
can
join.
Let's
see
I.
C
To
hear
that
this
one
helps
okay.
Well,
let
me
know
if
you
feel
better
yeah.
E
E
C
Can
also
ask
folks
who's
very
interested
in
joining
participating
today,
if
no
one
is
and
you're
not
feeling.
Well,
then
they
canceled
the
meeting.
Then.