►
From YouTube: CNCF Serverless WG Meeting - 2019-08-15
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
A
Alright
3
after
the
arrow
that's
going
to
get
started.
Kids
were
moko
later
I,
don't
pom-pom
I,
think
nothing
oops!
Sorry,
my
window
I,
don't
think
anything's
changed
with
the
AIS
and
stuff.
You
can
skip
that
community
time.
Is
there
anything
from
the
community
people
like
to
bring
up?
That's
not
on
the
agenda.
A
From
to
ever
mention,
Herman
Dobie
and
a
friend
of
mine
appointed
me
to
this
Accenture
project.
That's
an
open-source
project,
that's
actually
using
cloud
events,
so
we
were
thinking
that
we
could
use
them
as
a
as
the
third
end
user.
However,
do
you
think
it
would
be
really
nice
if
we
can
actually
get
a
little
more
than
just
three
I
think
in
the
background,
a
pinko
people
and
they're
working
on
it?
A
So
what
I'd
like
to
do
is
we
go
with
the
assumption
that
we
will
get
at
least
a
couple
more
before
we
actually
present
to
the
COC?
But
what
I'd
like
to
do
is
to
see
if
you
guys
feel
comfortable
with
us
actually
asking
to
get
on
a
future
toc
call
I
can't
imagine
it's
going
to
happen
for
a
couple
weeks
anyway,
because
they're,
usually
backlogged,
but
I
want
to
start
that
process.
Now.
B
A
Okay,
I
will
go
ahead
and
gas
to
get
up
on
the
agenda.
Obviously,
if
you
guys
review
the
slides
and
you
want
to
make
changes,
you
know
you're
more
than
welcome
to
make
suggestions
I'm
open
to
anything.
There
is,
as
I
said.
If
you
have
more
end
users,
you
want
to
list.
Please
do
there's
also
a
page
in
there
there's
a
it
talks
about
people
that
have
actually
adopted
the
spec
itself.
A
How
most
these
guys
are
not
end
users,
but
if
you
want
your
company
listed
here,
please
let
me
know-
and
we
can
add
it
just
to
to
show
that
people
are
actually
adopting
it
from
an
implementation
perspective,
not
from
end-user
perspective
just
to
help
it
along
okay,
so
sleep
put
on
the
agenda,
please
review
it
when
you
have
not
done
so
yet
will
get
some
last-minute
changes
in
there.
A
A
That's
the
guest,
the
minute
I
find
out
an
estimate
or
something
I'll.
Let
you
guys
know
there
Thanks
yep
all
right,
moving
forward,
coupe
con,
so
the
due
date
for
submitting
the
proposal
for
our
sessions
is
tomorrow.
So
I
need
input
from
you
guys.
We
have
a
choice.
We
can
either
do
an
intro
and
deep
dive
for
both
of
our
groups
or
we
can
do
a
90-minute
session
for
either
one
of
them.
A
My
current
thinking
is
that
doing
a
90
minute
session
for
the
server
this
one,
mainly
because
we
don't
have
that
much
material
and
I
actually
contemplated
dropping
down
to
just
a
135
minute
session,
but
I.
What
I'd
like
to
do
is
turn
it
into
a
birds
of
a
feather
session,
like
we
did
in
Barcelona,
where
we
basically
start
engaging
with
the
audience
to
find
out
what
their
status
is
relative
to
service
adoption,
in
particular,
if
they're
not
using
it.
Why
aren't
they
using
it?
You
know
what
are
the
roadblocks?
A
Is
it
just
fear
and
uncertainty
doubt
or
is
it
something
else
in
terms
of
lacking
functionality
electing
to
land
that
kind
of
stuff,
I
I
think
we
had
a
fairly
good
session?
We
did
it's
in
Barcelona,
even
though
we
had
a
relatively
small
audience.
I
I
do
think
it
might
be
valuable,
but
if
there
are
people
who
think
that
it
didn't
go
well
in
Barcelona
and
you
want
to
drop
down
to
35
minutes
and
just
talk
about
status
of
the
working
group
and
potential
future
things
I'm
open
to
that
as
well.
A
I
do
summary
of
spec
status
plans,
do
the
demo
Airport
demo
and
then
any
of
the
topic
you
guys
want
to
bring
up
in
there.
Just
do
it
all
in
one,
rather
than
trying
to
square
it
across
to
and
run
the
risk
of,
duplicating
things
between
the
two
sessions.
Anyway,
that's
my
current
thought
process
here
for
those
of
you
who
have
been
involved
in
the
past
or
want
to
be
about
in
the
future.
What
do
you
guys
think
about
this?
A
D
E
G
A
A
H
E
A
Okay,
so
I
guess
Scott.
What
you're
proposing
is
one
35
minute
session
that
just
gives
a
overview
of
the
project
itself
and
status
and
stuff,
and
then
in
production,
which
is
not
to
say
a
deep
dive.
I
mean
it
cut,
it
is,
but
it's
more
business-related,
I
guess
or
abuse
is
related.
What
do
people
think
about
that?.
A
E
A
I
A
I
J
We
went
through
this
exercise
for
naps,
we're
actually
going
to
combine
it
because
we
do
have
a
lot
of
material,
but
we're
gonna.
Do
it
in
a
way
where
we've
got
a
short
intro
and
then
we
kind
of
describe
a
use
case
and
how
we
approach
something,
and
then
the
second
half
is
going
down
into
the
real
details
into
code
into
configuration
that
sort
of
stuff.
So
people
can
leave
halfway
through
if
they
want.
A
A
What
look?
Let's
try
that
I
suspect
if
for
some
reason
as
we
work
it
through,
if
it
turns
out
ninety
minutes,
doesn't
work,
we
could
probably
convince
the
organizers
to
let
us
split
into
two
and
even
just
take
that
90
minutes
and
split
into
two
and
put
a
20-minute
block
in
between
them
for
people
to
walk
around
so
okay,
let's
do
that.
A
So
this
will
be
part,
one
in
part
2
and
we'll
figure
out
the
exact
details
later
I'm,
a
pink
some
of
you
guys
offline
later
today
to
get
a
pool
for
the
abstract
because
want
to
make
sure
I
summarized
what
we're
gonna
be
talking
about
appropriately,
so
keep
an
eye
out
for
that
all
right,
okay!
So
here's
the
current
proposal,
you
guys!
Okay
with
that.
A
A
A
A
A
A
Okay,
so
this
is
just
adding
some
guidance
that
says,
while
it
can
be
really
really
short,
it's
recommended
that
it
be
an
absolute
URI,
okay
and
we
made
it
so
that
it
has
to
be
a
non
empty.
Uri
reference
and
I
did
that
for
both
source
as
well,
and
what
was
the
other
one
schema
URL,
which
is
now
data
schema,
so
I
need
to
merge.
C
So
it's
so
the
distinction
between
absolute
and
relative
of
your
references,
it's
crystal
clear
and
that's
fine
I,
think
the
discussion
about
length
is,
does
not
add
value.
You
know
if
I'm
using
a
relative
URI
reference,
you
know
images
slash
cat
is
a
perfectly
good.
Fine,
relative,
URI
reference
I,
just
don't
think
blant
is.
A
C
A
B
C
A
A
A
Any
objections,
well,
this
room,
that's
what
the
restraint
section
is
so
yeah,
yeah
yeah
we're
not
completely
crystal
clear.
Sometimes
we
put
constraints
up
here
and
sometimes
do
it
down
there.
It's
a
little
a
little
fuzzy
on
that,
but
we
can
worry
about
that
later.
If
we
want
to
okay,
is
there
any
objection
to
me,
making
this
edit
or
I
consider
to
be
strictly
editorial?
A
Is
there
any
objection
to
making
this
editorial
change
after
the
call
and
assume
that
they
PR
still
approved
I'll
wait
till
we
get
one
or
two
LTE
teams
to
make
sure
I
didn't
with
any
kind
of
syntax
typo?
Well,
you
guys,
okay
with
me
merging
this
offline
I'm,
fine
with
it,
okay,
cool
any
objection.
Did
you
lay
down
that
path?
A
There,
no
pointer
and
I
read
me
to
the
SDK
thought
that
we
have
so
I
just
basically
added
that
the
only
other
thing
I
did
is
I
added
the
word
requirements
there.
The
title
of
doc
was
just
cloud
events
SDK
and
it
seemed
like
it
was
missing
something
or
some
other
word
after
that,
describing
what
it
was
so
I
put
the
word
requirements
there,
if
you
guys
have
another
word
like
recommendations
or
guidance
or
something
I'm,
okay,
with
switching
it
I
just
felt
like
not
having
another
word
in.
A
A
Okay
and
obviously,
if
you
guys
have
another
choice
of
words
in
there,
we
can
do
it
OPR
later.
It's
obviously
not
a
big
deal.
Thank
you
guys.
Alright,
next
biggie
Oh
tim
is
your
hand
still
up
where's,
that
old,
goose,
okay,
all
right,
so
hopefully
most
people
in
following
to
the
chats
that
are
going
on
to
the
PRS
and
stuff.
A
So
you
haven't
based
upon
the
conversations
that
were
going
back
and
forth
between
Christophe
ten
Clemens,
mainly
around
the
PR
that
they
christophe
submitted
as
a
wording
change
to
to
Clemens
I
made
a
suggestion
somewhere
in
one
of
those
comments
of
possibly
search
for
reducing
what
we
say
down
to
a
bare
minimum.
Clemens
suggested
I
turned
into
a
PR,
so
that's
what
I
basically
did,
and
so
this
PR
basically
does
a
couple
of
things.
A
First,
is
it
removes
the
notion
of
data
as
an
attribute
which
and
I
know
that,
technically,
when
we
do
this
structured
format,
it
appears
as
a
sibling
in
the,
for
example,
in
the
Jason
as
another
attribute,
but
from
the
abstract
perspective.
If
we
stop
referring
it
to
it
as
an
attribute,
then
we
don't
have
to
jump
through
hoops
to
say
our
data
types
apply
to
everything,
but
data
right
because
we
want
to
say
they
apply
to
all
attributes,
but
not
data.
A
It's
really
kind
of
funky,
so
I
remove
the
notion
of
data
as
an
attribute
and
just
said
just
I,
just
called
it
data
whenever
appropriate.
I
think
I
got
all
the
places
where
we
have
phrases
like
data
attribute
and
stuff
like
that.
I
may
have
missed
one
or
two,
so
if
you
guys
spot
them
in
the
future,
let
me
know,
but
that
was
the
that's.
The
first
big
change
I
made
is
stop
referring
to
data
as
an
attribute.
A
A
It
seemed
to
flow
nicer
to
just
say
we
define
attributes
and
here's
what
it
can
be
defined
as
in
terms
of
a
data
type,
and
it
just
seemed
like
it
flowed
a
little
better
I,
don't
think
I
necessarily
changed
any
of
the
semantics
in
here.
The
only
thing
I
did
do
is
added
boolean,
mainly
because
I
thought
that
would
that
should
be
a
valid
type
that
people
want
to
perhaps
define
for
an
extension.
It
seemed
like
a
very
natural
thing,
people
to
say
yes
or
no
on
or
off
kind
of
a
thing.
A
I
also
added
this
stuff
here
that
said,
all
attributes
must
be
a
scalar
type.
I,
don't
think
that
actually
changes
the
semantics.
Now
that
we've
killed
off
map
I
just
added
this,
to
make
it
explicitly
clear
that
they
must
be
a
scalar
and
they
must
not
be
of
complex
type,
meaning
maps
and
structures.
Since
we
explicitly
disallow
that
the
reason
I
added
this
was
because
we
now
don't
have
map
in
there
anymore.
A
But
we
don't
explicitly
say
you
can't
use
map
and
so
that
left
open
the
possibility
of
someone
actually
defining
an
extension
that
might
have
used
map
and
I
thought.
Based
upon
previous
discussions,
we
were
headed
on
the
path
of
banning
that
so
I
want
to
make
it
explicitly
clear,
but
then
the
bulk
of
the
change-
oh
another
one
in
the
optional
section
option
on
attribute
section,
I,
said
optional
or
conditional
because
as
Christoph
was
pointing
out
while
some
of
these
well
technically
all
of
these
are
optional.
A
They
actually
may
be
required
based
upon
what
you're,
adding
into
the
cloud
event.
So,
for
example,
the
content
data
encoding,
I'm,
sorry,
data
content
encoding
may
actually
be
required,
depending
on
your
civilization.
So
I
had
work
conditional
there
since
I
think
that
might
be
a
little
more
accurate,
but
the
real
folk
with
a
change
is
up
here
and
I'll.
Let
you
guys
read
this.
A
Basically,
all
I'm
trying
to
say
in
here
is
if,
when
you
serialize
data
and
you're
in
the
structured
format,
if
this
it's
a
normal
civilization
of
data,
does
not
align
with
the
serialization
of
your
envelope
itself,
you
may
need
to
base64
encode
it
and
in
particular
use
the.
Where
is
it
oh
yeah?
Well,
guess
this
is
on
the
one.
A
A
I
should
say
sorry
and
I
think
here's
the
last
big
section
now
in
data
we
remove
its
type.
So
it's
not
even
any
anymore.
It's
just
almost
undefined
I
actually
originally
had
the
word
undefined
in
there,
but
then
I
removed
it
/
Clements
suggestion
just
not
even
getting
any
confusion
about
it.
So
basically
just
says
it's
the
payload
defined
by
the
data
content.
Type
schemer
URL
points
to
it
talks
again
about
the
native
syntax.
A
If
you
need
it,
you're
supposed
to
use
a
data
content
encoding,
if
you,
if
the
native
syntax
per
get
a
cut
thing
encoding
or
its
native
type,
it
doesn't
align
with
your
envelope
and
therefore
you
must
use
the
data
content.
Encoding.
Ask
you
anyway:
yes
got
a
64
all
right,
I
said.
Hopefully
most
you
guys
weren't
following
the
conversation
during
the
week,
but
any
questions
on
that.
I
L
L
Right
now
sense
that
on
words
using
Jason
because
of
Jason's
ambiguity
in
in
terms
of
how
dates
are
encoded,
because
sometimes
they're
RFC
1123
and
sometimes
they're,
there
RFC
333
9,
which
means,
when
is
I,
so
the
other
one
is
not.
You
might
end
up
in
a
place
that
you
receive
it.
You
receive
the
day
as
Jason,
and
then
it's
you.
You
end
up
with
the
wrong
expectation
effectively
and
that's
the
only
that's
the
only
data
type
that
I'm
really
worried
about,
because
Jason
has
the
weakest
type
system
and.
L
L
So
is
that
something
comments
you
think
we
could
work
on
in
a
follow
up
here.
Yeah
I
just
want
to
point
that
point.
I
just
wanted
to
point
that
out
like
this.
That's
that's
a
rule.
I
would
love
to
add
to
the
Jason
event
format,
even
though
it
should
stay
out
of
the
data
thing.
I.
Think
taking
a
stance
while
Jason
doesn't
is
good.
L
A
M
A
A
A
A
Why
you're
think
about
that
Clements,
Roberto
I
assume
this?
Is
the
de'lon
you're
worried
about
right
I,
don't
think
that
line
is
necessarily
a
problem
per
se,
because
it?
This
is
just
saying
that
if
you
have
binary
data,
you
may
need
to
be
a
64
encoded,
because
it's
technically
possible
your
binary
data
could
be
the
word
hello,
which
does
not
technically
need
to
be
basically
for
encoded,
I,
see
where
you're
going
right.
H
A
That's
that's
my
rationale
for
why
I
think
it's
technically!
Okay,
if
you
think
that
sentence
is
gonna
cause
a
problem.
I'd
almost
rather
just
remove
the
entire
sentence,
since
it's
just
a
for
example
anyway,
but
I
thought
it
might
be
useful,
but
I
think
that
I
don't
have
a
huge
preference
on
that.
One.
No.
L
L
A
Holy
cow,
okay,
so
question
for
you
guys
and
thank
you
guys
very
much
especially
good
evening,
Clements
Kristoff
and
James,
hopefully
you're
listening
the
record
I
tell
you
guys
very
much
for
all
your
work
on
this
I
believe
that
allows
us
to
close
this
P
R
from
James
Wright
make
sure
we
look
at
the
right
one
right,
Christoph
and
Clements.
We
can
close
this
one
now
right.
N
A
Now
question:
for
you
guys,
there
are
two
issues
open
up
by
James
I
believe
that
we
cannot
close
both
of
them,
but
I
wanted
to
get
your
guys
take
on
it.
Just
before
I
did
that.
Can
we
close
this
one
as
a
result
of
all
the
various
pubes
that
we've
done
recently
between
killing
off
map
and
the
one
we
just
approved.
A
A
Cool
okay
I'll
go
with
the
assumption
we
can
meet
always
reopened
if
I'm,
wrong.
Okay,
cool
all
right
next,
thank
you
guys
very
much.
Let's
go
to
Tim's
PR
now,
I!
Don't
think
this
one's
ready
to
go
for
two
reasons:
one
I
missed
the
Tuesday
deadline.
However,
I
think
there
might
be
some
outstanding
questions
on
this
one.
So
Tim
you
want
to
talk
to
this
one
and
explain.
C
Yeah,
the
core
idea
was
this
thing:
you
should
say:
printable
Unicode
characters
and
there
ain't
no
such
thing.
Having
said
that,
the
sentiment
was
that
having
garbage
Unicode
characters
that
have
no
interoperability
value
and
no
semantics
is
a
bad
thing
and
I
did
a
hasty
first
step
of
that
just
ruling
up
the
control
characters.
But
if
people
buy
into
the
notion
that
we
would
like
to
be
fussy
about
the
repertoire
of
unicode
characters,
that's
allowed
in
the
right
term
non
data
attributes.
C
Then
we
should
develop
a
thoughtful
list
which
we
can
get
by
stealing
from
various
other
standards.
You
know
no
new
line,
characters,
no
non
character,
characters,
no
control
characters,
probably
know
naked
surrogates.
There's
a
well
known
list
of
Unicode
garbage
that
is
reasonably
straightforward
to
to
exclude.
N
L
A
C
A
A
Nevermind
that
commented
on
it,
so
basically
what
what
Evan
did
here
is
he
basically
produced
a
whole
bunch
of
different
sample
inputs
for
really
really
weird
cloud.
Events,
in
particular
jump
to
some
funky
stuff,
I,
think
more,
the
data
one
did
it
didn't
see
yeah,
so
he
did
some
weird
stuff
here
like
for
the
ID.
He
has
some
funky
characters
in
there.
Here's
a
better
example
right
he's
it's
a
whole
bunch
of
really
weird
stuff
now
quickly
wrong,
but
but
Tim
your
PR.
C
A
A
Agree:
okay,
so
anyway,
so
this
PR
is
basically
giving
a
whole
bunch
of
funky
cloud
events.
So
people
I,
guess
I
said
that's
Scots.
They
can
do
some
sort
of
performance,
testing
and
stuff,
but
make
sure
people
are
aware,
but
they
may
get
some
weird
characters
in
there
and
they
need
to
be
prepared
for
it
since
I'm,
not
sure
me,
people
have
actually
looked
at
this,
oh
and
we
need
to
wait
for
tens
PR
anyway.
Let
me
just
say
that
I'm
going
to
ask
people
to
please
review
it.
A
A
A
A
You
Europeans
tend
to
do
that
a
lot
yeah
all
right,
let's
see
technically
we're
done
with
all
the
PR
stuff.
So
let's
jump
to
a
discussion
item
that
Mark
suggested
last
week
that
we
have,
which
is
a
quick
discussion
or
maybe
not
clear,
but
a
discussion
around
our
governance
model
in
particular
around
the
way
we
do
voting
rights.
A
As
you
guys
all
know,
we
do
voting
rights
based
on
attendance
on
this
phone
call,
not
based
upon
PRS
and
stuff
like
that
or
can't
pee
our
captain
stuff,
which
is
slightly
different
than
some
open
source
projects
or
most
open
source
projects.
I
should
say:
I
will
state,
though,
that
from
my
experience
anyway,
I
think
Clemens.
You
may
experience
this
as
well
from
four
normal
standards
bodies
that
work
on
specs
a
lot
of
times.
A
A
So
that's
why
PR
cat
isn't
I
say
the
best
guy,
since
we
do
a
lot
of
collaborative
offline
discussions
and
back
and
forth
and
stuff
Issy
wordsmithing.
So
anyway,
do
people
feel
like
we
need
to
change
our
governance
model
or
you
know,
have
any
questions,
concerns
desire
to
for
changes
now
the
time
to
raise
it.
A
I
have
a
comment,
but,
as
anyone
else
want
to
comment
on
that
first,
okay
yeah,
my
reaction
to
that
is,
it
is
definitely
true
from
a
technically
voting
right
perspective.
Somebody
like
James,
who
I
believe
is
in
Australia,
he's
probably
never
gonna
give
boarding
rights.
However,
the
way
we
tend
to
operate
is
more
along
the
lines
of
almost
everybody
has
a
veto
option
right,
so
James
had
had
some
serious
concerns
about
that
part
of
this
point.
You
know
that
data
part
of
the
spec
you
open
to
PR.
A
We
went
back
and
forth
Clemens
bent
over
backwards
to
try
to
work
with
him.
It's
from
a
timezone
perspective.
So
in
my
opinion,
almost
anybody
tends
to
have
a
veto
power
in
terms
of
blocking
the
PRI
from
going
in.
Obviously,
at
some
point
we
do
end
up
taking
votes
at
that
point.
It's
a
shame
that
James
didn't
get
a
vote
necessarily
if
this
thing
axe
did
coming
out
the
formal
vote,
but
relative
to
PRS
getting
rushed
through
and
and
and
the
voting
members
get
the
steamroll
anybody.
A
B
A
O
A
Know
that
is
third
point
for
you
and,
like
I,
said
I'm.
Ok,
if
people
want
to
to
to
stay
there
here
in
the
chat
thing
and
avoid
the
head
count,
thing
I
don't
mind
getting
rid
of
that
I.
Just
to
be
honest,
the
reason
I
do
the
head
count.
Thing
is
because-
and
this
is
my
paranoia
based
upon
real
life
experience-
is
people
play
games
and
I've
had
cases
in
the
past
with
other
groups?
Not
this
one.
That's
actually
that's
not
sure
this
group
has
done
it
once
or
twice.
A
I
just
haven't
called
it
out,
but
in
previous
groups,
I've
been
in
people
will
do
things
like
add
their
name
to
the
agenda,
but
then
never
show
up.
If
all
you're
doing
is
letting
people
self
register
or
they'll
go
get
on
the
phone
call
and
their
name
will
appear
in
the
Xoom,
but
they
they
walk
away
permanently
right,
so
they
just
log
in
and
walk
away
and
that's
not
to
me
I'm
the
phone
call
now
granted.
A
You
know
once
you
guys
say
I'm
here
you
could
technically
vanish
and
I'll
never
know,
but
at
least
that
tells
me
you
were
at
least
awake
enough
to
actually
say
yes,
Doug
I'm.
Here,
that's
that's
the
only
reason
I
do
it.
I
mean
it's
just
to
try
to
avoid
game
playing
because
I
have
seen
it
done
quite
badly
in
the
past
and,
like
I
said
I've
seen
it
I
think
twice
so
far.
This
group
I
just
haven't
called
it
out.
That's.
A
D
A
A
A
B
A
The
reason
I
did
that
was
because
I
wanted
it
to
be
a
little
more
public
if
you
guys
are
comfortable
with
someone
mentioning
it
through
the
public
slack
Channel
I.
Just
don't
want
someone
to
ping
me
alone,
I'd
rather
more
than
one
person
and
see
it.
If
you
guys
want
me
to
add
to
the
public
slack
Channel
I'm,
okay
with
that
as
well.
A
A
A
Okay,
cool
I
will
make
that
change
and
I
will
wait
until
you
have
one
or
two
LG
teams
to
make
sure
I
didn't
mess
up
the
text,
how
they
make
that
change
all
right.
In
that
case,
believe
it
or
not,
we
are
actually
at
the
end
of
the
agenda.
Are
there?
Is
there
any
other
business
people
like
to
bring
up
at
all
I.
A
F
A
P
A
A
Okay,
we
need
to
decide
on
this
one
from
Thomas
from
ages
ago
to
me
to
see
whether
we've
actually
done
basically
testing
and
usage
testing
to
make
sure
we
actually
use
this
her
role
to
experience
this
one
is
Evans,
PR
I
think
we
can
close
this
one
I
think
Thomas,
I'm,
sorry
I
think
Tim's
will
deal
with
this
one
that
one
we
did
all
right.
Consider
moving
the
webhook
speck
that
one's
an
open
question
I,
think
Clemons.
You
have
to
comment
on
that.
A
I
need
to
look
at
your
comments
and
respond
if
any
worst
case
scenario.
If
we
actually
do
that,
it's
removing
a
document,
so
it
that'd
be
an
easy
thing.
If
we
do
well
worst
case,
is
it
but
best
case
scenarios
we
change
nothing.
So
do
you
think
I'm
very
bullying,
kind
of
a
thing
in
my
opinion,
so
yeah
I
think
I'm
almost
done
with
issues
I'm.
L
Looking
for
whether
we
there's
a
good
place
for
that
document,
yeah
we
haven't,
we
haven't
finished
a
discussion
yet
so
there's,
but
that's
also
going
to
be
interesting
then,
because
it's
like
there's
a
play,
I
think
there's
a
place
within
within
the
linux
foundation,
where
that
might
be
able
to
move
within
C
and
C
F
weather
might
be
able
to
move,
there's
a
suspect,
spec
team
and
then
there's
also
an
effort
within
w3c.
But
then
you
know
the
document
has
been
created
here.
So
there's
all
kinds
of
interesting
complications
on
moving
that
out.
L
The
w3c
I,
don't
even
know
how
that
works.
There
were
were
there's
some
some
Microsoft
Standards
people
who
talk
about
all
that
might
work
out
so
because
I'd
certainly
don't
have
been
in
the
spec
per
se.
Mm-Hmm
all
right.
You
know
I,
think
I,
think
the
mechanisms
that
we
have
there
are
useful,
specifically
the
abuse,
protection
and
and
just
a
the
fact
that
you
deliver
an
event
with
using
a
post.
That's
something
that
the
our
HTTP
binding
doesn't
specify.
L
A
Well:
okay,
yeah!
We
can
continue
our
discussion
offline,
but
I
think
that
might
be
the
only
one.
That's
actually
outstanding
Christophe,
but
yeah
I
think
I
think
over
the
last
couple
weeks
we
did
some
really
good
progress
here
on
closing
things
out
and
it
may
be.
You
know
we
may
get
to
the
point
where,
assuming
we
merged
Tim's
PR
next
week,
where
next
week
we
may
actually
consider
going
to
0.9
and
making
that
a
release
candidate
and
put
us
in
that.
You
know
testing
mode.
A
N
A
So
maybe
we
may
be
really
close
when
that
be
cool
if
we
went
1.0
before
koukaon
awesome
anyway,
don't
to
get
our
hopes
up
all
right,
any
other
topics,
then
all
right
cool.
In
that
case,
we
are
done.
Thank
you
guys
and
dis
reminder
and
four
minutes
we
have
the
SDK
call
if
you
guys
are
interested
in
that,
for
those
of
you
are
interested
I
should
say
all
right
cool.
Thank
you
guys.
Let's
talk
to
you
next
week,
all.