►
From YouTube: WG Component Standard Office Hours 20200901
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
Okay,
good
morning,
everyone
welcome
to
the
tuesday
september
1st
working
group
component
standard
office
hours.
This
is
our
hour
for
folks
to
drop
in
and
ask
questions
and
get
help
with
what
they
are
working
on.
I
am
mike
taufen
and
I'll
be
in
the
channel
today.
Answering
questions.
B
A
The
doc
looks
really
good.
I
think
your
edits
were
great.
I
had
a
few
comments
on
it,
but
overall,
I
think
it
looks
great.
I
think
so,
since
this
is
mostly
like
a
revision
of
the
existing
cap,
I
want
to
try
and
reuse
the
same
pull
request
if
possible.
A
So
I'm
trying
to
give
you
access
to
my
branch
that
backs
that
pull
request
so
that
you
can
just
push
over
it
with
the
updates.
This
is
the
first
time
I've
ever
tried
to
do
this
on
github.
So
let
me
know
if
you
got
like
you
should
get
like
an
invite
or
something
to
collaborate
on
my
fork
of
the
enhancements
repo
and
then
I
have
to
figure
out
how
to
make
sure
you
have
access
to
that
specific
branch.
B
Okay,
should
I
push
the
same
comment
like
as
question
comment.
A
You
can,
I
would
just
I
would
I
would
pull
pull
that
branch
down.
First,
add
a
new
commit
on
top
and
then
push
with
just
the
new
commit
just
to
start
just
to
see
if
that
works.
But
I
think
you
have
to
accept
the
invite
first
yeah,
that's
something:
okay,
great.
A
Yeah
so
I
pulled
down
mtof
and
so,
like
add,
add
my
reposa
remote,
pull
down
instant,
specific
branch
from
that
remote
or
just
get
fetch
the
remote
and
then
check
out
the
branch.
A
Yeah,
let's
talk
I
mean
we
can
talk
through
the
doc.
Do
you
have?
I
know,
there's
some
to-do's
in
there
that
I
thought
were
good.
Do
you
have
any
like
specific
questions.
B
A
Great
okay,
yep
yeah,
I
totally
got
what
you
were
doing
there
see,
I
think
definitely
yeah
like
we
can
add
some
diagrams.
I
thought
maybe
like
especially
when
we're
showing
how
complicated
the
case-by-case
merge
is.
It
would
be
worth
like
a
sequence
diagram
to
show
all
the
like
stuff
that
has
to
happen
in
that,
or
even
maybe
not
a
sequence
diagram,
maybe
just
like
a
list
of
all
the
steps
just
so
people
can
see
without
looking
at
the
example
like
oh
yeah,
that
is
pretty
complicated.
A
Up
to
you,
okay,
I
just
want
to
like
make
an
impression
on
people
like
yeah.
It
was
really
complicated.
We
probably
shouldn't
do
it.
We
can
offer
complicating
troubles
more.
A
And
then
I
thought
we
could
probably
come
up
with
a
few
more
specific
examples
or
reasons
that
show
why
the
strategic
merge
patch
is
simpler,
probably
actually
just
adding.
That
list
is
enough.
If
you
just
say
like
look
look
how
many
steps
there
are,
if
we
don't
do
strategic,
merge
patch
and
look
how
few
there
are.
If
we
do,
that's
probably
actually
enough
sorry.
A
So
in
there's
in
the
design
details
section,
there's
a
configuration
merge
example
and
one
of
the
paragraphs
sort
of
said,
like
this
simple
you
know,
shows
how
strategic
merge
patch
simplifies
the
implementation.
A
And
the
reason
it
gave
was
well,
it
already
uses
an
already
proven
method
from
api
machinery,
but
I
think
we
can
be
more
specific
about
that.
A
So,
like
I
added
one
that
says
like
we
avoid,
you
know
complexity
around
like
if
you
know
the
kind
that
we
implemented
the
case-by-case
merge
against,
isn't
the
same
kind
that
gets
read
in
that
an
automatic
conversion
might
happen,
and
that
might
screw
us
up
right
and
we
could
maybe
like
list
a
few
of
the
other
edge
cases,
but
also
like
just
show
how
it's
you
know
like
one
step
to
just:
do
the
merge
and
then
on
marshall.
A
A
A
Okay
yeah,
but
this
looks
like
pretty
good
overall,
I
just
like
clean
up
the
last
few
comments,
maybe
add
a
couple
diagrams
and
then
see
if
you
can
update
the
existing
cap.
B
A
And
then,
but
yes,
then,
for
the
example,
pr
that
has
the
strategic
merge
patch
example
in
it.
We
should
also
include
a
link
to
that
in
this
cap,
and
we
should
update
they
just
comment
on
that
pr
once
the
cap
is
up.
So
the
folks
who
commented
on
that
know
that
it's
available.
A
B
Them
I
I
have
like
I
put
it
to
do
in
the
example
more
like
what
corner
cases
are
we
missing
here?
You
know
yeah
what
can
really
screw
up
this.
This
game,
like.
A
Yeah
a
couple
things
that,
like
it's
possibilities
that
come
to
mind
so
this
wouldn't
this
wouldn't
be
like
technically
wrong,
but
it
might
be
confusing
to
users
if
the
merge
strategy
changes
from
flags
to
the
instance
specific,
like
strategic,
merge
patch.
A
That
said,
I
think
we
won.
I
commented
somewhere
on
here
that
I
think
it's
like
an
open
question.
I
think
we
want
to.
We
want
to
think
carefully
about
what
the
merge
strategy
should
be
for
each
one
of
these
options,
some
so,
for
example,
today
with
flags
the
only
parameter,
that's
a
true
merge
is
feature
gates
and
all
the
other
flags,
even
if
they
are
flags
that
specify
map-like
structures.
A
For
instance,
like
the
eviction
thresholds,
like
the
entire
eviction
threshold
map
will
be
replaced
by
whatever
you
set
in
flags,
whereas
with
feature
gates.
B
The
good
parts
that
we
can
control
this
kind
of
merges
in
the
in
the
tanks.
So
whatever
we
decide,
we
append
the
patch
strategy
for
it.
The
tag
in
this
struct,
like
patch
strategy,
merge
or
patch
track
strategy,
replace.
A
Right,
but
we
have
to
make
the
decision
and
that's
the
yeah,
that's
what
we
have
to
do.
The
other
thing
I
was
thinking
is
is
like
it
would
be
nice
if
user.
You
know
we
might
decide
someday
like
actually
it's
better.
If
users
can
specify
the
merge
strategy
that
they
want
to
use
because,
like
they're,
the
ones
who
are
actually
running
the
configuration.
A
A
Okay,
I
think
what
else.
A
I
think
we
need
to
make
sure
that
there's
a
requirement
in
here
that
the
both
the
config
and
instant
specific
config
files
are
a
only
specifying
one
object,
each
at
least
based
on
this
implementation
and
b
only
specifying
objects
that
have
the
same
api
group
and
api
version
so
that
we
don't
need
to
so
the
first
one
is,
so
we
don't
need
to
map
between
multiple
objects,
which
would
make
this
more
complicated
and
the
second
one
is
so
that
we
don't
need
to
do
conversions,
because
conversions
require
unmarshalling
and
we're
trying
to
do
this.
A
You
think
this
can
be
a
blocker
or
no,
I
think
it's
just
we
just
need
to
to.
I
think
we
should
just
explicitly
note
it.
I
don't.
I
don't
think
anyone
will
really
object
to
it.
Okay,.
B
A
B
B
B
A
Okay,
I
left
a
comment
with
some
thoughts
but
yeah
and
the
only
other
things
I
can
think
about
are
like
we
have
to
make
sure
we
test
it
really
well
and
that
we're
getting
the
expected
merge
results
for
all
these
parameters.
B
A
A
We
can
also
look
at
the
kinds
of
merge
strategies
that
are
used
on
other
api
objects
and
and
see
if
there
are
similarities
to
the
cubelets
configuration.
A
A
A
B
Dude,
but
what
was
the
deal?
That's
some
people
don't
want
to
use
this
totally
because
they
cannot
override
this
thing.
A
A
A
They,
I
think,
have
instant
specific
fields
actually
in
the
api
and
they
need
to
remove
those,
and
I
think
it's
I
think
it's
better
if
we
just
use
your
merge
approach
and
just
like
move
to
that,
like
it
doesn't
matter
what
is
in
your
api
because
you
can
set
it
from
either
place.
B
Yeah
from
from
the
conversation
on
the
scheduler
they
they
are
not
to
use
this.
A
I
don't
know
which
specific
conversation
but
like
we're
eventually
just
going
to
push
every
component
to
use
this
as
its
configuration
api.
Some
some
components
have
used
have
called
like
called
it
component
config
when
they
have
other
config
files
that,
like
don't,
represent
the
set
of
options
for
their
flags
and
they
use
those
for
other
features,
and
so
some
people
might
be
thinking
like.
Oh,
we
don't
have
a
new
feature,
we're
building
that,
like
loads,
a
new
config
from
a
file.
So
we
don't
need
this.
A
So
like
api
server
has
a
feature
where
you
can
configure
how
web
hooks
authenticate
to
web
hook
servers
via
like
a
file,
that's
local
to
the
api
server's
file
system
and
that
file
is
nominally
implemented
in
a
component.
Config
style.
A
So
it's
kind
of
like
a
different
use
case
for
component
config
versus
just
specifying
the
like.
You
know
basic
options
for
a
component,
but
yeah
we're
trying
to
move
all
of
them
in
that
direction.
A
Api
server
is
kind
of
like
the
most
sensitive
component.
So
there's
generally
a
lot
of
resistance
to
big
changes.
There.
B
A
A
Cool
so
see
you
next
week,
man
yeah,
you
too.
Let
me
know
if
you
have
any
questions.
Let
me
know
if
you
have
trouble
accessing
that
branch.