►
From YouTube: Kubernetes WG Apply 20180911
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
B
B
A
C
A
A
A
A
A
A
Figure
out
types
to
present
the
emerging
and
different
code,
I
think
I
actually
did
that
and
that's
also
my
conclusion
was
to
get
around
like
packet
boundaries
and
venturing
concerns.
I'm
just
gonna
write
a
little
type.
That's
convertible
to
the
unstructured
look.
You
can
consume
stuff
in
the
map.
String
interface
form.
So
if
you
look
in
that
PR,
which
I
will
put
a
link
to
it
also
will
deserialize
to
Yambol
and
back
without
losing
order.
So
that
should
be
convenient.
Okay,.
A
B
C
A
A
A
A
We
have
not
gotten
a
DI
review
yet
so
I
have
some
I
have
one
new
task
which
could
be
done
by
a
buddy.
I
was
going
to
do
it,
but
if
anyone
could
do
it,
which
is
the
the
repo
by
the
way
antwa
did
you
or
Jenny
approved
the
first
PR
I
made
to
our
new
repo,
like
fixing
the
like
fixing
the
readme
and
stuff?
A
C
A
A
C
A
The
fills
new
code,
which
is
what
we're
doing
right
now,
and
what
do
you
think
about
instead
of
using
that,
like
literally
take
whatever
cute
control
used
to
be
doing
and
whatever
strategic
merch
patch
used
to
be
doing
and
smash
them
together
into
one
directory.
So
we
get
behavior.
That's
exactly
the
same
as
today's.
B
A
There's
three
tracks
that
we
could
go
to
get
a
functioning
apply.
We
could
so
the
day
what
happens
is
cute
control?
Does
a
diff
constructs
a
patch
sends
it
to
the
server
you
merge
badge
applies
it.
That's
that's
one
path
by
which
apply
functions
and
that's
the
only
path
by
which
it
actually
functions.
There's
also
the
the
code
that
Phil
started
or
wrote
which,
as
far
as
I
know,
is
not
actually
battle
tested.
A
Yeah
is
that
true,
yeah,
so,
and
and
and
we
moved
it
once
and
I'd
like
to
move
it
again
and
of
course
the
third
path
is,
is
to
just
finish
the
reference
implementation
that
I
started
eventually
would
want
to
do
this
up
fast
right,
I,
I.
Think
I
want
to
do
the
third
path
anyway,
for
reasons
like
being
able
to
write
down
a
very
clear,
specific
age
clock,
so
other
people
can
can
repeat
what
we've
done
so
I'm
wondering
if
it
makes
sense
to
use,
fills
rewrite
since
it's
not
battle
tested.
A
But
if
we
don't
use
it,
then
we
have
to
figure
out
how
to
get
how
to
allocate
extract
the
both
the
diffing
code
from
queue,
control
and
the
strategic
merge
code
from
api
server
and
like
latch
them
all
together
and
I'm
fine
if
we
copy
stuff,
because
I
really
wanted
to
be
like
fixed
in
time.
So
we
can
always
like
use
it
as
a
as
a
comparison
point.
A
C
A
B
A
A
Mean
modulo
the
bugs
that
we'd
hope
to
fix,
but
I
I'm
wondering
if
the
bugs
that
we
have
to
fix
are
maybe
at
least
a
known
quantity
and
and
if
we
use
the
like
new
implementation,
that
bill
has
like
I,
don't
know
how
early
it's
been
tested
and
maybe
it's
got
maybe
a
fix.
Those
bugs,
but
maybe
as
a
bunch
of
other
bugs
I,
don't
know.
A
B
C
A
A
Don't
know
about
that,
I
think
I
think
no
matter
what
we
want
to
do.
We
need
the
we
need
the
API
change
that
we've
made
and
we
need
the
changes
to
the
patch
handler
that
we've
made.
We
like
all
the
stuff
that
we've
done
in
the
future
branch,
no
matter
what
we're
gonna
do.
We
need
all
those
changes.
Yeah,
that's.
C
A
C
C
A
B
A
So
and
like
I
also
like,
even
if
we
do
nothing
else
from
this
project,
but
take
the
code
out
of
cute
control
and
put
it
in
the
server,
that's
still
a
great
improvement
on
the
status
quo,
because,
true
because
right
now
you
cannot
use
this
code
from
any
language
from
from
any
like
library
right,
you
can't
call
it.
Every
client
has
to
either
implement
it
or
shell
out
or
do
something
different,
so
totally
good
enough
to
be
another
500
Shanta.
Yes,.
A
B
C
C
C
Yeah,
so
I
keep
the
cool
as
it
is
today
we
can
move
stuff
around
if
needed,
but
I
would
keep
the
code
as
it
is.
Instead
of
trying
to
move
SMP,
we
don't
know.
What's
what
that
means
and
what's
gonna
happen
as
we
do
that,
and
we
know
in
the
in
the
final
step.
We
don't
want
to
keep
these
Korean
anyway,
so
it'd
be
kind
of
weird
to
create
a
single
version
so
that
later
we
have
the
subversion
of
the
merging
code,
I.
A
C
A
A
A
B
C
B
C
No
just
apply
path:
okay
you'll
apply,
so
you
can
now
create
a
lot
date
on
the
apply
path
and
in
those
I
think
it's
telling
it
I'm
not
even
sure
what
insurance
that,
if
you
do
a
create
path,
it's
not
gonna
set
it,
and
this
one
should
be
easy.
I
think
what
you
are
supposed
to
do,
but
you
can
create
through
the
apply
verb
now
you
can
create
led
apply,
but
we
would
have
to
look.
C
A
B
C
Kevin,
do
you
mind
if
I
don't
even
think
about
it?
Okay
and
that's.
A
A
C
Yeah
I
mean
the
way
apply,
is
implementing
it
keeps
it
yell.
It
means
it's
not
going
to
be
I
know.
I'm
Who
am
I
kidding,
because
the
client
is
super.
It's
super
easy
in
the
client.
You
use
the
new
endpoint
right
because
there's
nothing
to
do
besides
sending
the
object,
so
it
should
be
super
easy.
Oh
you
think
it's
super
easy
right.
I
did
it
for
I
didn't
even
find
it.
We
thought
when
you
wanted
to
apply,
you
had
to
do
all
these
stands,
create
the
patch
and
the
patch
brother
blasts.