►
From YouTube: Kubernetes SIG Docs monthly APAC meeting for 20200826
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,
we're
recording
hello.
Everyone
today
is
tuesday
august
25th
2020,
and
this
is
the
weekly
meeting
of
kubernetes
dax.
Today
is
the
fourth
tuesday,
so
we
are
meeting
at
6
00
p.m.
Pacific
time
in
order
to
make
it
easier
for
folks
on
the
asia
pacific
region
to
attend.
A
So
it's
always
nice
to
see
us
yokosong
and
sujay
pillai
are
you?
Are
you
joining
us
from
from
somewhere
in
the
the
absolutely
huge
region
of
asia.
B
Yes,
I'm
from
currently
based
out
of
malaysia
penang
so
telling
is
a
mix
of
mainland
and
a
small
island
which
is
on
the
north
side
of
malaysia.
I
would
say
so
five
hours
from
the
capital
street
awesome
well,
thank
you
for
joining
us,
we're
glad
you're
here
yep.
I
just
once
joined
during
your
like
my
night
time
previously,
that's
where,
but
it
was
kind
of
like
very
late.
For
me.
A
Cool,
well,
that's
good
that
we
started
with
new
contributors.
A
A
Cj,
are
you
is
kubernetes
part
of
your
day
job,
or
are
you
just
here
because
you
love
the
community
until
the
time.
B
B
So
we
had
so
the
community
is
not
big
as
compared
to
other
community
that
I
have
seen
so
I
got
an
attendance
of
maximum
20
peoples,
excellent.
A
Well,
welcome
yeah
thanks.
So
let's
do
updates
and
reminders.
So
this
week's
pr
wrangler
it
was
scheduled
to
be
barney
aka
makos
cafe,
but
I
have
not
heard-
or
I
haven't
heard
from
barney
you're,
seeing
any
activity
from
him
for
a
while.
So
I
can
sort
of
wrangle
this
week
and
I
spent
like
yesterday
wrangling,
I
think,
a
fair
number.
A
I
can't
give
it
like
the
full
attention
that
it
deserves,
but
I
can
sort
of
wrangle
this
week,
but
it
would
be
a
good
idea
to
reach
out
to
barney
and
see
what
his
intentions
are
about
continuing
as
an
approver.
If
he
wants
to
stay
great,
if
not,
let's
make
sure
that
we
clean
up
the
owner's
files.
A
A
Awesome,
thank
you
cool,
I'm
looking
ahead
to
see
in
a
month
we
have
vineeth
also
is
scheduled,
and
it's
been
a
long
time
since
I
have
seen
activity
from
vineeth,
so
it
would
be
good
to
like
reach
out
to
him
proactively.
A
I
would
like
to
not
be
the
one
to
reach
out
this
time.
If,
if
someone
wants
to
reach
out
to
barney
and
just
say
hello
and
see,
what's
up
with
him,
that
would
be
awesome.
D
I'm
happy
to
take
that
and
I
can
reach
out
to
barney
and
beneath
not
a
problem.
Awesome.
A
Let's
move
on
to
the
agenda,
so
the
release
is
upon
us.
The
repo
is
frozen
so
right
now,
no
merges
in
the
in
the
repository
at
all,
and
thankfully
proud
now
has
functionality
that
makes
that
automatic.
So
like
don't
be
surprised
or
alarmed.
If
you're
seeing
things
get
approved
and
not
merge,
that's
normal.
It's
we.
A
We
freeze
the
repo
to
make
sure
that
no
last-minute
breaking
changes
are
introduced
and
that
we
can
make
a
clean
tag
at
the
end
of
the
current
release
and
the
start
of
the
next.
So
as
soon
as
we
have
the
tags
in
place
and
the
official
release
goes
out,
the
door
we
will
file
the
repo
and
a
whole
cascade
of
prs
will
merge
all
at
once.
A
There's
a
load-bearing
tide
blocker
in
in
that
issue.
So
anything
else
to
note
about
the
release
in
terms
of
like
details
and
how
it.
D
This
is
the
first
time
that
we're
testing
not
having
the
release
branch
until
later
in
the
release
cycle,
but
it's
a
little
different
for
this
time
around,
because
we
created
the
release
branch
once
like
early
on
and
then
created
everything
and
then
deleted
it,
and
so
we
recreated
it
and
everything
looked
good
as
far
as
just
for
context
and
awareness
when
I'm
talking
about
the
release
branch,
it's
the
branch
where
118
is
in
master
now
where
118
goes
to
b
after
we
do
the
release
for
119.,
but
so
the
the
branch
itself
was
created
early
on
configured,
we
deleted
it,
we
recreated
it
later.
D
A
Sure
it
seems,
though,
so
just
to
recap
really
quickly
you're
talking
specifically
about
the
the
the
named
branch
release
dash
1.18.
A
We
have
waited
until
much
later
in
the
cycle
like
we
did
at
the
the
we
created
the
first
we
said
wait.
This
is
a
terrible
idea.
We've
we've
had
problems
when
we
do
it.
This
way
we
deleted
it,
and
so
has
the
release
dash.
1.18
branch
been
created
again.
D
A
Okay
is,
and
just
to
double
check
is,
do
you
know
if
release
1.18
is
current
to
to
master.
D
Yep,
as
far
as
I
know,
yeah,
it's
looking
all
right.
A
Okay,
awesome
cool.
Thank
you.
It
seems,
like
we've,
had
a
lot
fewer
headaches
of
mispointed
prs
this
time
around
as
a
result
of
changing
how
we
do
that
so
yeah,
a
nice
process
improvement,
blog
issues,
anything
going
on
with
the
blog.
A
Looks
looks
good
all
right
issues
and
pr's.
So,
as
I
was
wrangling
prs,
I
was
looking
through
the
the
backlog
of
things
that
we
have
open
and
there
were
some
specific
pr's
that
have
been
hanging
out
for
a
while,
and
I
wanted
to
bring
them
up
in
community
and
like
work
our
way
through
them.
A
So,
let's
start
with
pr
for
client
libraries,
21
509..
This
is
taylor's
pr
aka
only
dole
taylor
is
the
release
lead
for
1.19.
So
I
expect
not
to
hear
or
see
from
him
for
like
two
weeks
like
all
of
this
week
and
then
next
week,
to
like
not
do
anything
and
just
remember
how
to
be
human,
so
is
there.
A
I
guess
I
would
ask
folks
to
take
a
look
at
this
and,
if
there's
anything
else
that
we
need
to
do
here,
then
let's
outline
that
clearly,
but
this
is
starting
to
become
a
a
pr
with
many
dependencies
stacking
up
behind
it.
So
I
don't
want
to
delay
this
too
much
longer
if
we
can
any
thoughts
about
like
what
to
do
with
this.
I
think
tim's
feedback
here
is
really
solid
about
how
to
label
and
handle
client
libraries,
but
I
think
celeste,
your
third-party
content
warning
is
also
gonna
play
into
this.
A
So
any
thoughts
here
about
how
to
proceed
with
this.
C
A
A
And
I
think
if
I,
if
I
read
tim's
feedback
correctly,
tim's
proposal
is
to
keep
it
and
slap
a
giant
warning
label
on
it,
saying
we
don't
maintain
this
we're
not
directly
responsible
for
this
content,
use
it
your
own
peril.
E
No,
I
I
would
prefer
we
keep
this
a
simple
cable,
because
it's
not
a
lot
of
content
to
to
maintain
is
anyway,
it's
a
community
work.
If
someone's
having
to
find
that
one
one
of
the
libraries
has
gone
bad
and
maintained
there
could
be
a
proposal
to
delete
it,
remove
it.
E
We
cannot
afford
that,
but
providing
a
link
for
traders
is
is
always
a
good
thing
that
that's
just
my
opinion.
C
E
E
Sorry,
then,
let
me
put
it
this
way.
Maybe
people
are
aware
of
the
operator
concept
today,
right
now,
when
I'm
developing
a
operator
using
a
language
other
than
go,
I
will
need
to
interact.
C
E
C
E
A
Okay,
so
I
don't
love
keeping
this,
but
it
sounds
like
there's
a
sufficient
use
case
to
keep
it.
I
do
not
want
to
get
in
the
habit
of
relitigating
the
cap
about
third-party
content
every
time
something
comes
up
like
this.
I
also
don't
want
to
get
into
the
habit
of
maintaining
outdated
content
and
having
people
open
issues
against
outdated
content.
I
also
don't
want
to
be
in
the
habit
of
constantly
dealing
with
vendors
trying
to
one-up
each
other
by
adding
themselves
to
this
list
and
deleting
each
other.
A
So
the
solution
that
I'm
willing
to
accept
on
this
is
to
slap
a
third-party
content,
warning
on
it
and
say
use
at
your
own
peril,
but
yeah.
I
I.
E
Yeah
yeah,
we
can
move
it,
yeah
move
it
to
say
a
reference
section
and
put
a
large
warning
there.
This
content
is
not.
C
Yeah
yeah,
and
I
think
I
think
we
have
to
accept
that
level
of
ambiguity,
as
we've
discovered-
and
I
think
tim
mentioned
in
another
issue
around
this-
that
it
might
be
time
to
pr
against
our
own
cap
and
update
the
cap
itself
and
say
like
these
are,
as
we
discover
a
separate.
A
C
C
D
A
This
is
why
I
don't
want
this
content
to
live
here.
I
would
say
that
is
a
separate
discussion.
The
way
that
we're
going
to
get
through
this
is
to
hyper
focus
on
the
the
context
of
this
pr
itself.
So
this
pr
in
particular
needs
to
rebase.
A
A
I'm
sorry
to
be
a
pain
about
this,
but
I
like,
I
gave
five
months
of
my
life
to
this
fight
and
I
don't
want
to
keep
having
this
fight
every
time
that
we
deal
with
third
party
content,
which
doesn't
make
me
necessarily
the
most
unbiased
arbiter
on
this
decision.
A
So
I
guess
what
I
want
to
do
here
is
make
sure
that
we
scope
the
discussion
to
this
particular
pr
and
say:
is
this
mergeable
if
these
conditions
are
met,
so
I
guess
like
quick,
quick
up
or
down
if
the
specific
feedback
is
back,
can
we
move
all
right?
That
looks.
That's
that
looks
clear.
So
let's
do
that
and
let's
move
on
to
the
next
pr.
A
Okay,
so
better
cka
experience,
two
two:
nine
seven,
three.
A
Jim,
I
wonder
if
you
can
maybe
recommend
a
way
forward
here,
because
this
contributor
opened
a
pr
that
looks
like
specifically
to
be
helpful
to
people
taking
the
cka
exam.
A
And
given
our
experience
with
given
what
tim
shared
about
his
own
experience
with
the
cka
exam,
I
suspect
this
particular
contributor
would
be
happy
just
having
this
pr
in
here,
whether
it's
open
or
closed,
because
it
makes
it
searchable,
and
anyone
could
find
this
result.
So,
but
I
wonder
jim,
if
you
can
like
recommend
more
about
how
to
proceed.
D
Yeah,
this
is
definitely
a
hairy
issue
for,
for
general
context,
for
folks
keeping
up
with
some
of
the
different
pieces
going
on
here,
neolit
or
lubamir
from
the
sig
club,
multi-cluster
life
cycle
or,
more
importantly,
cube
adm
from
sig
multi-cluster
life
cycle
ended
up,
creating
a
pr
removing
some
of
the
cni
install
instructions.
D
The
reason
being
was
that
there
was,
I
think,
three
examples.
I
believe
it
was
psyllium
calico
and
maybe
flanat
flannel
was
the
other
one
and
what
happened
was
everybody
was
coming
in
and
saying
add
my
cni
to
this
list
as
well,
and
it's
really
not
cubay
dm's
position
to
say
this
is
one
winner
of
the
cni's.
Here's
the
you
know,
loser
cni's,
it's
really
just
saying
in
the
kubernetes
community,
you
can
choose
your
own
cni.
You
can
run
with
whatever
cni
you
want.
D
Here's
a
link
to
that
and
it'd
be
an
external
link
to
a
plethora
of
of
container
networking
interfaces.
The
the
challenges,
the
cka
exam
explicitly
allows
you
to
use
kubernetes
documentation
and
not
external
links.
So
anybody
who
is
leveraging
that
documentation
for
configuring,
a
cluster
or
anything
of
that
nature
couldn't
actually
click
on
those
links
to
get
the
install
instructions
for
the
cni's
which
are
now
hosted
externally.
D
So
the
pr
was
in
an
effort
to
just
kind
of
not
play
favorites,
which
I
think
was
in
the
right.
You
know
spirit
as
well
as
in
line
with
the
cap
that
we've
been
talking
about
about
third-party
content,
so
I
think
the
pr
overall
was
a
net
positive,
but
now
there's
that
issue
of
how
do
you
resolve
it
from
an
exam
taker's
perspective,
eliminating
this
content
moving
it
to
an
outside
resource
which
is
not
accessible
by
by
anybody
here.
So
as
far
as
a
way
forward,
I'm
I
don't
have
a
great
recommendation.
D
My
my
gut
would
say
potentially
just
have
calico,
be
the
default,
the
de
facto
networking,
the
cni
there.
The
reason
I
say
that
is,
I
know
that
there's
a
lot
of
infrastructure
and
testing
that
is
done
through
kubernetes
and
ci
signal
as
well,
and
a
lot
of
those
clusters
are
running
either
kind
or
clusters
that
are
running
calico
cni.
D
It
seems
like
a
lot
of
I
guess.
People
are
getting
started
with
kubernetes
start
with
calico
as
kind
of
the
de
facto
standard,
maybe
canal,
which
is
calicon
flannel
mixed
together,
it's
kind
of
a
hard
spot
to
be,
but
instead
of
saying
we're
picking,
favorites
it's
just
going
down
to
a
single
provider.
That
is
one
of
the
more
popular
ones.
That's
been
one
of
the
most,
I
would
say
I
don't
say
most
maintained,
but
one
of
the
longest.
You
know
battle
tested
cni's
that
have
been
out
there.
D
I
haven't
looked
at
crans's
pr
what
what
actually
was
being
implemented
when
I
was
scrolling
through
there,
it
looked
like
there
was
some
flannel
components
as
well,
so
the
analytics
helium
also
got
added
back
it.
So
it
looks
like
what
what
he
ultimately
did
was
add
all
of
the
cni's
back
to
the
original
page
here
under
cluster
administration.
Networking.
A
In
229.73,
it
also
looks
like
he
added
some
quick
install
instructions
so.
D
D
Thing
about
the
pr
with
the
quick
install
instructions
is,
it
does
keep
in
the
spirit
of
having
many
cni's.
It's
not
just
saying
here's
the
one
but
you're
going
to
have
the
separate
problem,
except
for
one
because
dm's
problem,
it
will
be
networking's
problem
on
this
page,
where
everybody
who
has
a
cni
will
want
to
be
represented
on
this
page,
which
maybe
that's
a
bridge
or
a
mean
to
the
end.
I
guess
of
solving
this
problem.
I'm
not
not
too
sure.
F
I
also
don't
like
have
a
question,
so
I
know
there's
a
ck
version
too
coming
out
in
september,
first
or
sometime
in
september,
which
does
still
include
bootstrapping
and
installing
in
kate's
cluster,
so
they
would
have
to
still
add
cni.
F
I
remember
someone
mentioning
that
they're
going
to
talk
with
the
cka,
the
people
in
charge
of
the
cka
to
have
the
one-liner
include
possibly
included
in
the
in
the
props.
I
I
have
taken
ck
and
c-cad,
and
sometimes
they
do
provide
something
that
needs
in
terms
of
a
command
or
because
they
will
provide
you
what's
needed.
Would
that
be
a
possibility.
A
I
think
it
absolutely
would
so
funny
enough
jim
and
I
had
a
conversation
with
the
cncf
folks
who
maintained
the
cka
and
ckad
exams
last
week
and
they're.
Ask
of
us
was
to
go
through
and
revise
the
content
this
week,
so
I
am
waiting
on
access
to
the
test
material
in
order
to
be
able
to
revise
it.
But
that
is
a
really
solid
suggestion.
A
I
think
jimmy
you.
I
seem
to
recall
you
making
basically
the
same
suggestion.
D
Yeah,
that
was
one
of
my
suggestions.
Early
on
is
saying
that
you
know
no
one's
picking
a
cni
winner,
loser
just
the
cka
exam
saying:
here's
the
instructions
needed
to
install
calico,
for
example,
the
rest
of
the
exam
would
still
be.
You
know
the
administration
portion
of
kubernetes,
but
having
it
just
be
listed,
there
is
the
apply
or
the
ammo
there.
So
they
don't
need
to
actually
go
to
a
resource
would
absolutely
be
a
solution
which,
which
I
recommended
and
then
I'm
not
sure
if
it
was
implemented.
D
A
Okay,
that
seems
to
me
like
a
much
better
way
forward
than
returning
to
a
place
where
we're
vetting
content
from
third
party.
A
Providers
any
other
thoughts
or
input
there.
A
No
that's
real
yeah.
This
can
keep
for
like
another
week
or
two.
I
think
I
mean
people
are
still
passing
the
cka
so
yeah.
Let's,
let's
table
this
for
two
weeks
and
hopefully
by
then
we
will
have
access
to
test
material
and
can
make
specific
recommendations
about
how
to
improve
that.
A
Okay,
let's
move
on
see
23083
for
moving
the
search
page.
This
seems
really
straightforward
to
me,
but
I
don't
want
to
approve
this
in
a
vacuum.
A
Basically,
making
our
site
search
page
play
nicely
with
doxy
and.
C
C
A
And
also
to
make
it
clear
that
the
search
results
page
has
like
the
banner
for
it
says,
specifically
search
results
and
not
documentation
so
that
it's
not
misleading.
In
that
sense,
it's
a
a
clearer
user
experience
about
what
exactly
someone's
looking
at,
and
that
alone
seems
like
a
worthwhile
reason
to.
C
I
need
to
test
this
a
bit
more,
but
I
agree
it
seems
reasonably
straightforward.
So
how
about
this?
If
you're
cool
with
it,
you
can
give
the
looks
good
and
then
I
will
go
through
and
assign
myself
and
do
like
a
thorough
give
this
like
a
thorough
poke
to
make
sure
that
we
aren't
gonna.
Do
anything
terrible.
A
Let's
see
I
mean
I
will
ping
you
for
the
approve
on
this
one.
A
A
Hey
cool,
thank
you,
so
sub
module
updates
jim.
I
saw
your
pr
23434
to
make
sure
that
we
are
pulling
in
the
latest
security
update
from
the
doxie
theme.
I
I
tested
the
deploy,
but
apart
from
just
running
your
commands
locally
and
everything
looking
good
to
me,
I
don't
know
how
like
what
else
we
would
do
to
test
this.
D
That's
pretty
much
the
exact
boat
that
I
was
in
it
deployed.
I
was
like
hey,
it
worked
and
I
noticed
that
the
actual
commit
hashes
were
changed
on
the
sub
module.
This
is
the
first
time
I've
ever
updated
a
sub
module
to
pull
it
into
the
main
remote.
So
I
thought
that
was
pretty
cool
from
just
a
general
learning
experience
but
yeah,
I
think,
kicking
the
tires
on
this
and
then
obviously
wait
for
the
release
to
settle
down,
and
maybe
we
merge
it.
A
Sounds
good.
The
only
question
I
had
was
the
looks
like
github
is
not
liking.
This
particular
commit
hash.
It
is
if
you
look
at
the
particular
status
for
what
is
it
for
bb
64.
A
7.
this
commit
does
not
belong
to
any
branch
on
this
repository
and
whether
that's
github
temporarily
like
losing
its
mind
or
something
in
particular,
so.
D
I
I
think,
and
please,
if
there's
someone
who
knows
more
about
git,
to
tell
me
where
I'm
wrong
here,
but
I'm
pretty
sure
that's
as
designed
if
you
were
to
go
into
kubernetes
website
with
pulling
the
sub-module
update.
If
you
then
change
directories
into
that
sub
module
and
do
like
git
status,
you're
going
to
see
a
detached
head
based
on
a
commit
hash.
Oh.
A
D
A
Yeah,
that's
git.
Sub
module
is
the
worst
get
mini
game
yeah,
but
I
I
suspect,
you're
right.
I
think
this
is
exactly
that
detached
head
behavior.
That
happens
when
you
update
a
sub
module.
D
So
the
other
question,
too,
that
is
worth
bringing
up
to
the
group,
is
so
this
is
the
security
update
that
I
referenced
was
something
about
click
jacking
and
some
way
that
the
default
template
was
set
up
to
allow
certain
activities
which
is
not
really
a
best
practice
or
recommendation
from
a
security
perspective.
D
When
we,
when
we
roll
over
master
to
the
previous
118
branch,
I
believe
that
that
release
branch
is
still
going
to
have
the
sub
module
version.
That
is
older.
Maybe
with
these
vulnerabilities
in
it
and
with
this
doxie
theme,
do
we
need
to
then
recursively
go
to
all
of
our
previous
releases
that
leverage
this
and
update.
C
So
can
somebody
I'm
unfamiliar
with
basically
what
happens
when
we
roll
a
release
for
documentation,
so
when
that
old
version
is
created,
what
has
happened
and
the
question
I
asked
this
is
because
this
is
a
static
site
generator
ergo.
Should
we
not
just
be
pulling
static
files
which
are
not
necessarily
subject
to
the
same
security
vulnerabilities
by
nature
of
being
static.
A
A
Manually
update
the
branch
with
a
submodule
when
a
submodule
rules
like
this.
I
think
that
we
have
to
manually
update
the
branch
to
pull
those
that
update
in
because
it
will
build
correctly,
but
the
changes
won't
live
in
the
repository.
They'll
only
live
in
the
deployment,
if
that
makes
sense.
A
So
the
way,
if
you
look
at
the
makefile
and
the
actual
build
command,
the
build
command
pulls
the
submodule.
It
was
a
git
submodule
update,
recursive
update
depth,
one
means
the
build
environment
will
always
have
those
changes,
but
the
repository
itself
won't
have
those
changes,
so
there
will
be
a
diff
rents
between
what
lives
in
the
repository
versus
what
lives
in
the
deployment.
If
we
don't
pull
those
changes
into
the
repository.
D
D
C
Every
time
we
release
a
new
version,
we
create
a
branch
for
the
old
version
and
we
deploy
from
that.
E
D
Yeah
exactly
so
yeah.
We
won't
have
this
issue
this
time,
but
we
will
have
it
once
118
is
on
release
118.
If
this
merges
after
the
release,
which
it
most
likely
will
given
the
the
current
state.
But
then,
as
we
move
forward
and
120
comes
out,
we
have
118
and
119
etcetera,
etcetera.
We
will
have
to
keep
in
mind
sub
module
updates
need
to
go
throughout
the
different
previous
releases
and
celeste
answer
your
question.
D
I
know
you're
thinking
a
lot
about
this
release
process
and
some
of
the
static
files,
and
I
completely
agree
I've.
I've
had
this
thought
as
well,
and
netlify
supports
this
kind
of
monorepo
thing
where
there's
multiple
pieces
of
a
repo,
all
sharing
different,
smaller
modules
of
a
repo,
I
haven't
been
able
to
wrap
my
head
around
it
in
a
functional
sense
that
is
worthy
of
completely.
D
You
know
edge
sketching
this
entire
process
that
has
been
kind
of
crafted
over
the
previous
releases,
but
if
there's
another
way
where,
instead
of
going
with
a
branch-based
model,
another
model,
that's
maintainable,
supportable
from
a
release
team
from
a
sick,
docs
team.
I
think
it's
worthy
of
at
least
discussing.
A
I
think
this
will
be
a
one-off
situation
in
that,
just
because
of
the
timing
of
merging
this
particular
pr,
because
the
branch,
the
pr
the
branch
is
sorry
the
repo
is
frozen.
So
the
repo
will
thaw
this
pr
will
merge
and
we'll
have
to
make
the
change
retroactively.
Even
though,
technically
it's
concurrent
during
the
118
release
yeah,
I
think
it's
just
the
timing
that
makes
this
particular
release.
A
Weird
and
yes,
there's
probably
a
a
way
to
mono
repo,
but
that's
like
a
whole
separate
conversation,
and
it's
like
quarterly
planning
conversation
so
yeah.
But
I
agree:
let's,
let's
always
keep
an
eye
out
for
a
better
way
to
do
things.
The
way
that
we're
doing
things
is
the
result
of
many
iterations
of
experiment,
and
just
because
it
works
doesn't
mean
it's
necessarily
the
best.
A
Okay,
any,
I
think,
that's
that's
what
we
have
on
the
agenda
for
pr's
any
other
particular
thorny
pr's
to
discuss.
We've
got
like
20
minutes
left,
so
we
got
some,
no
all
good.
Okay,
let's
move
on.
So
one
thing
I
wanted
to
make
folks
aware
of
is
on
the
14th
of
september.
A
Well,
okay,
so
the
cncf
we
are
hosting
four
technical
writers
for
google
season
of
docs.
That's
a
program
that
google
runs
to
introduce
technical
writers
to
open
source
contribution
and
the
their
three
writers
are
on
other
cncf
projects
vote.
One
writer
is
dedicated
specifically
to
improving
the
way
that
we
serve
generated
api
references
and
it
is
philippe
martin
who
is
a
long-standing
contributor
already
for
the
french
localization
and
also
in
sig
api
machinery.
A
So
basically,
what
felipe
is
going
to
do
over
three
months,
beginning
on
september,
14th
is
take
advantage
of
doxy's
built-in
functionality
to
serve
nicely
rendered
generated
apis
in
a
way
that
isn't
just
the
the
the
cumbersome
way
that
we
generate
and
serve
api's
references.
Now
I'm
trying
to
is
there
a
way
to
make
this
short,
so
the
kubernetes
api
collection
of
apis
is
all
defined
in
an
open
api
script
and
felipe
is
going
to
our
spec.
A
Sorry
and
philippe
is
going
to
figure
out
a
way
to
import
those
specs
whenever
they
update
in
the
and
they
live
in
the
kubernetes
kubernetes
repo,
whatever
they
update
felipe,
is
going
to
find
a
way
to
make
sure
that
that
is
automatically
ports
to
the
website
repo
and
that
the
website
repo
rebuilds,
so
rather
than
only
generating
api
references
once
a
quarter
during
the
release
process,
the
goal
is
to
make
it
a
continuous
delivery
and
to
also
make
the
output
pretty
and
more
browsable.
A
So
that
is
the
thing
that's
happening.
It's
awesome.
This
has
been
my
end
game
for,
like
three
years,
is
better
api
documentation,
a
better
api
reference
experience,
so
you
may
see
some
pr's
feel
free
to
review
and
add
and
comment.
A
Another
thing
we
have
a
persian
language
localization
in
progress,
looks
like
they're
getting
close.
I
just
wanted
to
boost
that
as
an
awesome
thing,
that's
happening,
we're
actually
getting
to
play
with
some
new
hugo
functionality
that
renders
languages
from
right
to
left,
and
so
it's
cool
that
hugo
has
that
built
in
style
guide
editions
in
the
course
of
reviewing
prs
over
the
past
two
weeks.
There
are
two
things
that
I
see
repeatedly
that
are
like
like
sand
in
an
oyster,
but
it's
gonna
be
an
ugly
pearl.
A
One
is
avoid
colloquial
language,
so
things
like
using
software
out
of
the
box
who
opens
a
box
to
get
software
anymore.
When
was
the
last
time
you
literally
opened
a
box
to
get
software
for
me,
it
was
like
1997..
A
It
like
it
just
doesn't
make
sense
anymore.
So
that's
a
colloquialism,
even
though
it's
common,
even
though
a
fairly
large
subset
of
people
know
what
that
means.
There's
a
fair
number
of
people
who
don't
and
how
do
you
localize
that?
So
I
would
like
to
update
our
style
guide
to
specifically
warn
against
colloquial
language,
and
I'm
haunted
by
the
idea
that
we
probably
already
have
that
in
the
style
guide.
A
We're
just
not
applying
it
enough
and
the
other
one
is
to
so
there's
a
best
practice
for
documentation
and
it's
whenever
you
use
an
acronym
on
a
page
or
an
abbreviation,
define
it
on
first
use.
So
if
I
were
to
say
cni,
I
would
say
a
container
network
interface
open
parenthesis,
cni
close
parenthesis,
so
you
save
it
full
term
first
and
then
give
the
abbreviation
for
it,
and
then
you
can
use
the
abbreviation
thereafter.
A
These
feel
fairly
non-controversial.
But
I
am
curious
and
happy
for
input
and
reflection
on
this.
C
C
I
mean
these:
are
these
are
pretty
standard?
One
thing
I
was
gonna
mention
to
the
people
on
this
call
is
divya
one
of
the
release
shadows
on
119.
A
All
right
so
celeste.
E
So
speaking
of
style
guide,
do
we
have
any
guide
any
guidance
or
instructions
on
using
the
thought
codes,
such
as
tooltip
glossary
things,
I'm
seeing
these
kind
of
circles
floating
everywhere
when
we
are
localizing
these
circles?
We
comment
a
paragraph
out,
but
the
circle
still
works
it's
kind
of
annoying,
but
do
we
have
any
guidance
on
how
and
where
to
use
them.
A
We
do
there
is,
they
are
in
the
contribution
guide.
We
have
sections
specifically
for
hugo
short
codes.
I
don't
think
that
they're
in
an
obvious
place,
so
I
think
there
it's
probably
there's
an
opportunity
there
to
reorganize
that
content
and
make
it
more
visible.
A
A
Okay,
cool
jim,
I
notice
you
have
a
note
about
localization
analytics
and
improvements.
D
Yeah
and
just
a
quick
comment
about
the
custom
hugo
short
codes,
I
think
I
found
the
page
that
you're
referencing
out
the
word
in
chat
there.
It
might
be,
might
be
worthwhile
to
update
how
to
localize
some
of
these
short
codes,
in
addition
to
just
general
use.
D
E
A
Interesting:
okay,
yeah.
That
sounds
like
a
hugo
specific
issue
where,
where
maybe
like
the
short
code,
is
overriding.
The
the
commenting
inside
of
inside
of
the
block.
E
Yeah,
the
solid
code
still
works.
If
it
is
a
html
comment,
if
it
it
may,
it
can
become
completed,
comes
out
if
it's
inside
a
markdown
comment
or
hugo
comments.
D
I
think
this
might
be
a
issue
with
the
moving
from
black
friday
to
goldmark,
and
I
believe
that
goldmark
doesn't
render
html
on
the
page
itself,
but
if
we
had
a
short
code
that
that
allowed
you
to
comment
out,
the
the
entire
section
would
probably
solve
this.
I
I
believe
I'll
have
to
to
see
about
what
this
would
look
like
to
implement.
But
when
you
have
html
on
the
page-
and
I
forget
what
exactly
happens-
they
render
basically
the
they
render
the
short
codes
before
they
render
the
content
on
page.
I
believe.
D
Cool
and
and
yeah
for
the
localization
and
analytics
improvements,
just
a
real,
quick
update.
I
sent
out
some
some
localization
pdfs
and
I
got
a
lot
of
great
feedback
and
I'm
trying
to
implement
a
better
dashboard
to
share
with
the
wider
localization
themes.
So
I
just
wanted
to.
Let
folks
know
that
I
haven't
forgot
about
that.
That
is
still
in
progress.
D
I'm
planning
on
evaluating
all
the
feedback,
creating
better
dashboards
and
then
potentially
adding
some
sort
of
either
additional
mailing
list
or
the
existing
mailing
list
to
send
those
out
and
folks
could
filter
them.
If
they
don't
want
them.
I
think
there's
a
lot
more
value
in
the
localization
teams
getting
it,
and
if
it
annoys
people
they
can
create
filters
to
avoid
them,
and
then
the
last
piece
here
is.
D
D
So
I
think
not
only
from
a
localization
standpoint
but
from
the
english
you
know
documentation
be
good
to
leverage
the.
Is
this
helpful
or
not
piece
there
and
from
where
I
was
looking?
It's
not
there's
not
a
lot
of
people
who
use
it
over
the
course
of
a
month.
I
would
say:
are
the
most
the
page
of
the
highest
amount
of
feedback
given
to
us
around
30
or
so
clicks
on
that
is
this
important,
or
is
this
helpful
yes
or
no,
but
nevertheless
it's
feedback?
It's
there.
I
don't
think
we
leverage
it.
A
One
quick
thought
there
there's
a
really
valuable
piece
of
documentation.
A
Wisdom
only
measure
things
you're
prepared
to
act
upon.
A
So
if
we're
not
acting
on
the
was
this
page
helpful,
I
then
I
would
say:
let's,
let's
remove
it
so
jim,
if
you
like,
if
you
can
find
a
clear
way
forward
to
implement
to
to
to
measure
and
act
on
this
feedback,
then
let's
keep
it.
But
if
it's
only
getting
like
30
clicks
a
month,
then
I
would
say
also
like,
let's
consider
removing
it.
A
I
I
don't
want
it
to
turn
into
a
time
sync.
If
it's
something
that
a
a
tiny
percentage
of
our
monthly
website
traffic
sees.
D
Well,
yeah
and
I'll
take
a
look
at
the
one
piece
that
might
be
more
beneficial
for
localization
teams.
Is
there
there's
not
a
ton
of
negative
feedback,
but
there
are
a
lot
of
no
positive
feedback
on
certain
localization
documents,
and
I
don't
know
whether
that
stages
aren't
localized
and
maybe
someone's
not
happy,
they're,
not
localized,
but
you
you
see
an
actual
cycle.
There
might
help
localization
teams,
but
once
again
I
completely
agree
if
we're
not
ready
to
act
on
it.
Let's
take
a
look
at
it.
A
So
if
only
30
people
are
clicking
a
month,
we
have
two
million
site
visits,
page
views
per
month,
two
hundred
thousand
twenty
thousand.
Two
thousand
two
hundred
twenty.
That's
point:
zero:
zero,
zero
one,
five
percent
of
our
web
traffic.
So
I
I
guess,
consider
that
advisedly.
B
D
C
C
D
Yeah
and
that's
kind
of
what
I
was
looking,
it's
kind
of
strange,
the
way
hugo
and
doxie
work-
and
I
didn't
know
any
of
this
stuff
until
digging
into
it
a
little
bit
but
there's
a
yes
or
no
button,
it
only
registers
yeses
and
if
you
click
no,
it
registers
it
as
an
event,
but
then
it's
a
percentage
of
overall.
D
So
if
everyone
clicks,
no,
you
have
a
zero
percent
success
of
that
page,
whereas
if
50
of
the
people
click
yes,
50
of
the
people,
click
no,
you
see
a
50
and
you
can
assume
that
it's
50
healthy,
positive,
happy
where
you
see
the
zero
percent
pages,
meaning
that
out
of
everyone
who
clicked
it.
Nobody
clicked
that
it
was
yes,
helpful
feedback.
D
You
see
that
a
lot
on
localization
pages,
where
there's
around
12
on
average
users
per
page,
where
they
said
that
nothing
on
this
page
is
helpful
without
looking
into
the
analytics
too
far
or
getting
too
deep
into
this
discovery,
it's
very
possible.
These
are
pages
that
haven't
been
translated
or
localized.
Yet
so,
if
I'm
a
non-english
speaker-
and
I
see
a
page
that
hasn't
been
translated
and
I'm
on
the
translated
branch,
I
would
very
quickly
click
no.
This
is
not
helpful
at
the
very
end
of
that
page.
A
Yeah,
I
know
that's
real
yeah,
fair
enough-
that
that
is
actually
a
really
interesting
pathway
to
follow
like
to
compare
the
the
percentage
of
happiness
on
a
page
versus
the
journey
that
brought
a
user
to
that
page.
A
Fair
anything
more
to
say
about
that.
A
Awesome,
thank
you
for
following
up
with
that
jim.
Thank
you
for
persisting.
There
celeste
editing
with
divya
yeah.
C
So
divya
reached
out
via
dm
to
ask
some
questions
about
how
we
edit
docs
and
what
our
standards
are
for
content
and
for
language,
and
so
I
often
do
a
calendly
link
to
have
an
informal
discussion
about
how
to
edit
the
docs.
But
this
has
actually
come
up
a
few
times
from
sig
release
related
people.
So
I
thought
I
would
bring
it
here.
C
Would
anybody
here
be
interested
in
a
workshop
on
how
to
edit
content
for
for
for
language
and
like
how
to
actually
improve
things
that
we
see
from
a
pure
language
perspective,
because
right
now
that
knowledge
lives
with
zach?
And
I
and
also
probably
tim
banister
in
there,
because
he's
british.
C
Yeah
and
it
looked
with
jennifer
rondo
as
well,
but
this
is
also
a
question
that
I
have
received
at
many
points
in
my
career
from
many
different
engineers
and
many
different
permutations.
So
if
there's
broader
interest
I'll
throw
out
a
doodle
to
find
a
time,
oh
god
about
editing.
A
Sorry,
no
it's
my
my
recoiling
is
specific
to
doodle
as
a
functionality,
not
to
the
idea
of
finding
a
time
that
were
that's
convenient
to
everyone,
so
jeffrey
klein.
I
know
that
you
are
a
technical
writer
by
trade.
I
wonder
if
this
would
be
of
interest
to
you
as
well.
B
C
Would
cool,
I
will
throw
a
doodle
in
sigdocs
I'll
float
it
over
sig
releases
way
and
yeah.
D
I
think
this
would
be
great
and
a
strong,
strong
plus
one
from
my
end
as
well
I'll,
be
happy
to
attend
something
like
that
for
sure,
and
I
was
thinking
as
we
brought
this
up.
D
I
wouldn't
be
opposed,
and
I'm
volunteering,
I
guess
so
feel,
free
to
tell
me
that
you
would
be
opposed,
but
I
wouldn't
be
opposed
to
even
sharing
tips
and
tricks
during
the
meeting
as
they
come
up.
Here's
content
editing
that
I've
done
where
you
know
here's
the
problem
here
is
the
outcome
of
you
know
technical
editor
hour.
A
Want
to
have
like
my
roy
batty
background,
I've
seen
things
you,
people
wouldn't
believe.
No,
that's
a
that's
a
really
good
idea,
yeah.
So
it's
the
last
excellent
idea
and
thank
you
also
for
sharpening
the
project
there.
C
A
Ignore
me,
I
I'm
an
old
man
uphill
in
the
snow,
both
ways
cool
also
so
like
celeste
and
jim,
and
I
have
taken
on
some
projects
to
do.
But
the
spirit
of
the
one
of
the
values
of
the
project
is
for
folks
who
are
new
step
up,
expect
to
step
up
and
expect
for
us
to
go
away.
So
if
there
is
something
that
you
are
interested
in
absolutely
jump
in,
do
not
wait
for
permission,
jump
in
throw
an
elbow
on
that
get
involved.
A
So
yeah
kubernetes
is
not
really
a
spectator.
Sport
come
on
in
cool.
We
have
four
minutes
left
anything
else
that
folks
want
to
talk
about.
In
the
time
we
have.
A
Oh
cool
thanks
everyone.
This
is
really
good.
I
got
really
good
participation
and
thanks
for
diving
in
deep
on
the
pr's
and
really
good
discussion
all
right.
Everyone
have
an
excellent
week,
enjoy
the
release
process.
May
it
succeed
quietly
and
be
uneventful.