►
From YouTube: WG Component Standard 20191022
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
All
right
good
morning,
everyone
welcome
to
the
tuesday
october
22nd
working
group
complaint
standard
meeting
alex
I'll.
Let
you
kick
it
off
since
you're.
First
on
the
agenda.
B
Yep
so
just
a
quick
update
regarding
the
lenient
or
like
strict
decoders
for
cubelet
there's
like
one
more
comment,
jordan
added
and
I
will
fix
that-
and
I
think
then
this
can
be
merged
from
what
I
gathered.
B
Good
and
following
up
our
conversation,
I
created
a
pull
request
for
lenient
path
for
scheduler
and
so
yeah.
I
submitted
that
already,
if
you,
if
you
want
to
have
a
look,
I
think
I
accidentally
as
well
and
the
other
one
so
basically
phoenix
joe
he's.
His
name
is
phoenix
blue
on
slack
he
was
the
one
who
like
during
the
mentorship
call.
B
He
put
his
name
down
the
strict
decoder
topic,
basically
so
reached
out
to
me
asking
how
he
can
you
know
help,
and
I,
like,
on
the
weekend,
spend
some
time
walking
through
my
work
and
explain
a
little
bit
about
things,
and
I
said
you
know
if
you
want
you
can
try
to
tackle
this
linear
path
for
the
eq
proxy.
It's
pretty
much
everything
you
need
is
in
those
previous
pr's
and
he
created
it,
and
I
had
like
a
comment
on.
B
If
you
guys
want
to
look
over
it,
I
think
this
this
might
be
a
good
place
to
get
him
started
on
the
on
the
work.
B
No
worries
happy
to
help
very
cool
yeah,
so
yeah,
if
you
yeah,
just
when
you
have
some
time,
look
for
those
pr's,
maybe
find
something
add
some
comments
and
I
hope
we
can.
A
C
B
A
good
way
forward,
you
know
if,
if
no
one
wants
to
take
this
or
contact
us,
I
would
love
to.
You
know
work
on
this.
Since
I've
been,
I
don't
know,
working
in
this
area,
I
guess
a
little
bit
now.
I
just
wanted
to
know
you
know
what
would
be
the
best
way
forward
how
to
approach
this.
If
you
have
any
advice,
write
a
cap.
A
Would
be
the
first
step
for
something
major
like
this
remember
try
to
include
like
historical
uses,
like
the
dummy
field,
that
I
guess
a
lot
of
qbm
people
are
using
for
customize.
B
A
C
So
actually,
I
tried
using
the
full
blown
meta
v1
object,
meta
in
the
cube
adm's
component
config,
basically
in
the
beta3,
which
is
coming
off
ready
cycle,
but
it
was
basically
a
total
mess
and
so
I'll
probably
go
with
strip
down
object
method
which
is
currently
stripped
down
to
just
the
name
field.
What
made
it
a
mess?
C
Well,
actually,
basically
meta,
v1
object.
Meta
is
a
couple
of
different
things
coupled
into
a
single
struct,
so
one
is
basically
the
core
definition
of
an
object
which
is
basically
a
name,
a
namespace
generated
name
and
like
these
kinds
of
fields,
and
the
rest
is
basically
the
representation
of
this
object
into
the
api
server.
So
this
is
basically
the
creation,
timestamp
deletion,
timestamp,
finalizers
and
stuff
like
that,
and
these
are
actually
the
fields
that
are
causing
most
problems.
C
Basically,
the
fuzzer
is
always
broken.
You
need
to
go
and
patch
it
quite
a
lot
to
make
it
work
and
all
the
time
you
were
serializing,
you
get
the
bunch
of
fields
like
creation,
time,
stump
and
stuff
like
that.
That's
basically
pointing
to
no,
because
these
are
entirely
meaningful,
like
don't
don't
mean
anything
for
a
component
config
and
they're
not
needed,
but
they're.
Simply
there.
B
So
when
you
say
better
use
the
stripped-down
version,
is
there
any
reference
on
that
or
how
would
that
actually
work.
C
Well,
I'm
in
the
process
of
putting
up
a
pr,
that's,
hopefully
going
to
land
like
being
submitted
actually
before
the
end
of
this
week
for
kubernetes
new
company
config,
but
there
is
also
some
work
that
was
done.
I
think
in
either
alpha
2
or
alpha
3
of
the
cluster
api.
C
So
the
folks
there
are
using
a
stripped-down
object,
meta
well,
not
as
straight
down
as
mine,
but
considerably
considerably
more
stripped
down
than
the
original
one
and
yeah
they
basically
narrowed
down
the
fields
to
just
the
ones
that
are
describing
the
object
and
nothing
api
server
specific
there.
C
Basically
going
the
object
method
way
is
like
stepping
on
the
landmine,
so
I
would
recommend
initially
to
go
and
talk
to
api
machinery
and
basically
try
to
tell
them
the
idea,
because
this
will
be
probably
the
most
difficult
part
of
it.
A
B
Would
that
would
be
like
just
you
know,
ping
them
in
slack
or
join
their
weekly
meeting
and
discuss
it
with
them
or
would
be
the
best
approach?
Maybe
just
once
you
have
a
cap,
make
sure
that
they're
copied
on
it.
C
C
Yeah,
basically,
that's
the
the
case.
Probably
anonymous
structures,
basically
except
the
name,
is
always
empty,
or
some
fixed
shrink
like
singleton,
something
like
that.
But
then
there
are
lots
of
different
cases.
For
example,
someone
may
wish
to
basically
slam
a
metadata
name
field
to
a,
for
example,
cluster
configuration
and
basically
describe
the
name
of
the
cluster
that
this
configuration
is
attached
to
by
the
this
name,
and
this
may
be
much
more
like
appealing
to
people
and
they
may
expect
it.
B
Okay,
cool
russ.
I
might
I
might
come
back
to
you
on
that
one.
I
might
have
some
questions
if
that's
okay.
B
No,
I
don't
I'll.
Just
like
you
guys
are
at
kubecon
north
america.
I
guess
yes,
I
will
be
yeah.
Is
there
anything
planned
for
component
centers
like
some
event
or
something.
A
B
B
Okay,
yeah,
I
was
just
wondering
if
you
know
I
think
I
mentioned
it
before.
If
you
could
do
something
for
kubecon
eu
still
a
bit
time
ahead,
it's
like
in
march
or
april.
I
think
just
to
you
know,
keep
it
on
the
agenda.
I
guess
you're
quite
easy.
A
A
Yeah,
that
would
be
cool.
Do
you
know
if
the
cfp's
gone
out
for
eu
yet.
B
Be
on
holiday
for
the
last
two
weeks
in
in
november,
so
I
just
wanted
to
like
put
it
on
the
agenda.
I
would
be
happy
to
to
do
something
or
to
help
or
something
like
that.
I
mean
given
a
couple
talks
before
you
know.
I'm
sure
you
can
figure
something
out
to
you
know
maybe
recruit
some
new.
B
Some
new
contributors.
A
Cool
sounds
good
all
right
thanks
so
much
everyone
remember
if
you
want
to
come
to
the
mentorship
meeting,
it's
at
10
a.m,
pacific!
So
that's
in
about
an
hour.