►
From YouTube: Kubernetes SIG CLI 20190911
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
A
A
Think
that
the
next
meeting
in
two
weeks
we
will
devote
to
117
planning,
and
so,
if
you
have
any
proposals,
items
that
you
want
to
have
done
enter
in
117
release
frame,
please
make
sure
to
add
them
either.
Most
probably
the
agenda
will
be
the
best
place
and
show
up
during
the
meeting
next
week,
and
we
will
discuss
the
major
themes
for
117.
A
A
B
A
A
We
will
also
go
a
little
deeper
on
the
topic
during
probably
the
next
week.
The
to
the
the
meeting
in
two
weeks
for
planning's
of
the
people
that
are
interested
in
in
helping
out,
maybe
with
cube
CTL
independence
can
also
find
a
piece
for
them
to
work.
On
a
quick
update
from
the
test
grid.
I
was
looking
at
the
testing
great
for
the
past
two
weeks.
It
was
mostly
stable.
We
had
some
glitches
after
the
PR
for
keep
CTO
logs
merged.
A
That
was
just
a
a
small
problem.
Another
one
was
that
there
were
some
glitches
in
the
fort
or
attached
commands
failing
two
days
ago
due
to
some
problems
with
GCB,
but
they're
everything
is
currently
stable,
so
there
should
be.
There
should
not
be
any
problems.
Also,
we've
added
cube
currently
six
CLI
mailing
list
to
two
tests
infra,
so
that
whenever
there
is
a
failure
we
were
with
regards
to
our
tests.
We
will
get
an
email
notification
to
our
mailing
list,
but
I
don't
expect
these
to
be
a
high
volume.
A
Emails
I
think
these
are
the
most
the
main
topics
now
I
would
like
to
move
to
the
topic
section
and
the
only
one
that
we
currently
have
is
a
feature
request
for
supporting
home
that
you
conflict
that
the
directory
is
I
was
looking
for
for
the
PR
author
Dave
Kerr
present
with
us.
I,
don't
see
him
on
the
list.
Is
the
person
that
added
the
topic
to
the
agenda
present
on
the
call
to
speak
up
about
this
yeah.
B
A
A
Okay,
so
a
reminder
that
in
two
weeks
we
will
be
doing
planning
I'll,
probably
send
an
additional
reminder
to
the
sick,
CLI
mailing
list,
so
that
people
are
aware.
I
know
that
this
accuse
ability
has
some
topics.
I
spoke
with
Valerie
over
Twitter
some
time
ago.
I
would
like
to
sync
with
the
cig
usability.
They
have
some
feedback
that
I
would
like
to
gather
and
I'll
start
addressing
some
of
the
things
during
the
117
release.
C
C
That
is
now
working,
and
maybe
one
question
to
you
and
do
you
have
any
infrastructure
for
tracking
downloads
from
github
releases?
So
I
do
know
that
there
is
a
total
download
count
that
you
can
get
from
get
up,
but
there's
no
history
for
the
number
of
downloads.
So
if
anybody
knows
a
tool
which
can
do
that,
that
would
be
great
I.
D
D
Kurt
Smith
at
Kui
could
serve
an
elegant
role
for
crew,
that
I
could
say,
crew
cast
with
a
KA
just
for
the
sake
of
argument,
install
Tecton,
for
example,
and
we
then
subsequently,
when
he
say,
coo
cuddle,
Tecton
view
my
pipeline.
You
would
get
a
window
popping
up
showing
you
a
visualization
of
your
electron
pipeline,
but
as
far
as
sort
of
consumption
is
concerned,
it
all
happens
mediated
by
the
crew.
Install
experience
under
the
covers
the
cask.
A
little
shim
will
be
the
proxy
for
downloading
the
Kui
component
tree
and
dealing
with
all
those
bits.
D
C
B
D
Wasn't
a
matter
of
having
a
UI
for
installing
plugins
you
just.
You
would
see
that
to
do
that
from
the
CLI
just
like
when
I
open,
homebrew
I
can
say,
brew,
cask,
install
Emacs
and
what
happen
that
the
result
is
that
I
get
an
app
that
goes.
Application
for
Emacs,
so
I
would
still
I.
Would
I
would
like
to
be
able
to
say,
for
example,
as
user
crew
cask
or
even
as
crew
install
Tecton
or
crew
cast
installed
Tech
time.
D
But
what
the
challenge
is
that
we
have
there's
a
every
UI
extension
to
coop
kernel
has
some
shared
componentry.
Oh
the
electron,
for
example,
bits
needed
to
pop
up
that
wouldn't
know
those
are
common
to
every
visual
extension
to
KU
cuddle,
and
so
when
I
say,
install
Tecton
I
just
want
to
have
that
core
components,
the
electron
bits
and
the
Tecton
plugin,
and
when
I
install
the
coop
kernel
extensions
and
on
install
the
Aquila
extension
and
so
on
as
I
as
a
user.
D
C
D
Essentially,
I
can
I
can
scope
out
it
on
the
next
Co
the
proposal,
but
it's
essentially
that
the
idea
that
that
crew
right
now,
just
like,
with
homebrew
on
its
own,
install
its
great
about
installing
ASCII
terminal
oriented
applications,
but
to
be
able
to
have
a
portal
into
an
application
form
native
visual
world.
That's
the
kind
of
thing
that
cask
offers
to
brew
and
we
could
offer
Co
we
can
offer
than
analog
for
a
crew
that
was
the
better
base.
So.
C
Maybe
you're
right
direction
to
go
with
this
is
so
if
this
is
just
a
set
of
special
plugins,
which
will
have
a
UI
component
which
will
be
used
through
cui
then
I
think
it
would
make
sense
to
wait
for
multiple
indexes
and
crew
and
then
create
a
special
GUI
index
where
all
the
UI
plugins
can
be
hosted.
Would
that
make
sense.
D
That
would
definitely
help
there's
another
there's
another
part
to
it,
which
is
that
if
every
plugin
has
its
own,
like
I,
said,
let's
say
of
a
Tecton
plugin
and
a
kiala
plug-in
just
for
the
sake
of
argument.
I
think
what
I'd
like
to
avoid
is
having
a
full
200
megabyte
chromium
instance
for
every
plugin
and
so
the
the
Cask.
D
Would
the
the
cuy
ask
cask
metaphor
would
give
us
that
that
that
proxy,
to
be
able
to
have
a
single
electron
instance
on
my
laptop
for
all
the
plugins
and
that
when
I
install
a
plugin
I'm,
just
getting
the
the
extra
bits
needed
to
to
manifest
that
plug-in.
So
that's
that's
the
architectural
advantage
of
having
a
cask
like
middleman
yeah.
A
D
A
See
any
problem
yeah,
that's
definitely
something
that
worth
exploring
I'm,
not
sure.
If
next
call
will
be
good
good
one
since,
like
I
mentioned
before,
the
next
call
will
be
devoted
to
117,
planning
and
I,
think
we
can
pretty
easily
fill
in
the
entire
the
entire
one
hour
that
we
have
slightly
less
weeks.
C
C
A
C
B
A
E
So
we
plan
to
deprecate
to
Petrus
fielding
customization
dot
llamo
fell,
so
the
reason
for
that
is
in
our
last
release,
we
added
the
extended
patch
which
can
support
both
strategic,
much
cash
and
Jason
patch,
so
that
the
type
of
the
patch
is
detect,
how
to
manipulate,
buy
customized
based
on
the
content,
so
user
don't
need
to
specify
whether
is
a
strategic
emerge,
patch
or
adjacent
patch,
and
the
extended
patch
can
also
be
applied
to
multiple
objects
based
on
a
selector.
That's
specified.
E
So
this
this
new
feature,
or
this
extended
patch,
can
cover
the
functionality
that
are
provided
by
current
two
fields
like
patches,
strategic,
merge
and
patches
Jason
6,
9,
O
2.
So
those
are
not
good
names,
so
we
are
thinking
about
to
deprecated
those
two
fields
since
now,
if
we
don't
deprecated
them,
we
have
three
fields
can
can
handle
the
pattern
and
one
of
them
can
cover
the
functionality
of
the
other
two.
So.
E
You
support
the
previous,
like
the
county
user.
With
those
fields
we
we
are
adding
a
fixed
command,
we
have
a
customized
fix
command
and
we
are
adding
some
changes
in
that
command
to
convert.
The
current
feel
those
two
fields
to
the
new
extended
patch
field,
so
user
just
need
to
call
customized
fix,
and
then
it
will
be
converted
to
the
new
phone.
They
would
donate
to
Mary
about
it,
and
that
is
one
thing
besides
that,
currently
we
are
writing
some
documentation
for
different
things.
One
is
about
writing
the
document
for
inland
patch.
E
A
B
So
just
a
quick
update,
I
think,
as
I
mentioned
before,
there's
three
three
of
the
commands
that
are
keeping
us
from
having
all
of
the
code
moved
into
staging
about.
95%
of
it
is
moved
into
staging
now
now
that
it's
117,
we
can
remove
one
of
those
commands
which
is
coop
cuddle
convert,
so
the
only
remaining
commands
that
are
that
are
keeping
us
that
have
dependencies
on
kubernetes
core
keeping
us
from
moving
everything
into
staging
is
coop
code
will
get
and
coo
cuddle,
auth
reconcile,
and
so
the
coop
code
will
get.
B
We
are
actually
very
close
to
I've
been
refactoring
printing
tests
in
order
to
be
able
to
pull
part
of
the
printing
into
our
CLI
runtime
and
once
that
happens,
I
think
we're
1
to
2
PRS
away
from
having
that
removed
and
then
finally,
the
auth
reconcile
dependency.
The
dependency
on
the
are
back
to
be
honest,
I'm,
not
exactly
sure
where
that
stands.
A
That's
it.
Okay,
I
have
one
question
with
regards
to
convert
I.
Remember
we
at
some
point
in
time
when
we
initially
deprecated
convert.
We
thought
about
the
idea
that
currently,
if
a
user
tries
to
invoke
the
convert
command,
we
will
inform
them
that
there
is
no
such
command
anymore
and
then,
if
they're
interested
in
they
need
to
instantly
plug.
In
that.
B
A
Exactly
because
I'm
not
sure
how
much
tension
will
be
around
us
entirely,
removing
the
convert
command,
I'm,
sure,
they're,
just
yeah,
sure,
yeah
I
think
at
least
for
one
seventh
for
one
release.
We
can
put
this
kind
of
information
and
that
it
was
deprecated,
it's
removed.
It
lives
in
a
separate
place
if
you're
still
interested
it.
This
is
how
you
get
it,
and
starting
from
118
bulges,
see
nothing
and
I.