►
From YouTube: Kubernetes Sig Docs 20180529
Description
Meeting notes: https://docs.google.com/document/d/1Ds87eRiNZeXwRBEbFr6Z7ukjbTow5RQcNZLaSvWWQsE/
The Kubernetes special interest group for documentation (SIG Docs) meets weekly to discuss improving Kubernetes documentation. This video is the meeting for 29 May 2018.
https://github.com/kubernetes/website
A
A
C
My
science
Messiah
yeah,
yes,
hi,
welcome,
yeah
hi,
I'm,
Messiah,
I'm,
writing,
serial
publication
and
books
for
kubernetes
and
I
developed
GK
like
container
platform
for
on-premise
and
I
organized
cloud
native
meetup,
Tokyo
and
after
pipe
at
in
the
contributor
summit
at
KU
we
become
and
first
I,
send
10
years
to
website.
Last
week,
yeah
I
would
like
to
review
and
I
see.
Zack
and
I
would
like
to
contribute
kubernetes.
So
today,
I
attend
here.
Awesome.
C
E
E
C
B
B
F
About
that
I
never
introduced
myself.
I
only
just
got
a
little
bit
familiar
at
writing
day
with
I
worked
closely
with
Steve
Perry
at
writing
day
at
the
Portland
conference
and
I'd
like
to
contribute,
but
I'm,
really
a
newbie,
so
I've
been
a
technical
writer
for
a
long
time,
but
I
want
to
become
familiar
with
open
source
and
contributing,
but
I
have
things
to
learn
still
so
yeah.
F
B
One
of
the
things
that
you
can
do
to
help
contribute
is
to
like
help
edit
documents,
because
a
lot
of
the
content
is
written
by
engineers
and
stuff.
So
you
know
you
can
help
them
with
structure,
and
you
know
language
things
like
that,
because
then
yeah,
because
then
that
should
help.
You
also
learn
the
concepts
as
well
right.
F
F
B
F
A
A
G
A
A
So,
as
you
think
about
it,
if
you
are
moved
to
be
the
Wrangler
for
next
week
feel
free
to
update
the
agenda
here
or
drop
a
line
into
the
Select
channel,
and
we
will
get
you
slated
in
and
queued
up
all
right.
Just
another
reminder
that
misty
is
the
release
Meister
for
1.11,
we
changed
hands
before
I
went
out
on
vacation,
misty
I
would
love
to
catch
up
with
you
about
how
that
has
gone.
Zak
I
would
like
to
check
in
with
you
and
see
how
you
are
following,
along
with
the
release
master
process.
D
D
So
we
still
have
seven
line
items
on
the
spreadsheet,
whose
back
state
is
unknown.
The
cigarettes,
meaning
is
directly
after
the
front,
so
I
will
be
letting
those
people
know
in
that
meeting
what
the
state
is
as
well.
I
did
go
through
last
Thursday
and
paying
all
of
the
github
issues
that
are
related
to
features
for
1.11
and
remind
people
about
that
deadline
and
from
that
painting
came
some
updates
that
some
creatures
would
not
make
it
into
111,
which
nobody
has
told
anybody
that
so
that
was
useful.
D
I
also
have
been
working
with
AI
and
Steve
and
Augustus
on
the
future
tracking
spreadsheet
itself.
I
gave
them
some
recommendations.
I
opened
a
github
issue
in
the
cig,
release
or
I.
Think
sync
planning
repo
I,
don't
remember
what
ever
the
repos
that
they
use
some
feature
requests
and
we
also
added
a
feature
sets
sheet
onto
that
spreadsheet.
So
that's
how
I
know
that
there
are
seven
issues
whose
dock
state
is
unknown.
Some
sort
of
general
planning
improvements
there
I
think
just
because
they're
fresh
eyes,
so
maybe
I
had
some
new
ideas.
D
We
are
able
to
merge
quite
a
few
hours.
I
did
a
rebase
last
week
of
the
one
on
eleven
branch,
and
things
are
going
pretty
well.
I
did
all
of
your
checklist
items
that
you
gave
me
before
you
left
and
I
feel
like
these
are
pretty
solid,
I'm
happy
to
get
feedback.
I
would
love
to
get
feedback
since
I'm
new
on
this.
I
I've
been
noticing
that
we
get
a
lot
of
PRS,
whereas
someone
fixes
a
few
broken
links,
and
you
know
we,
we
appreciate
those
but
I-
think
it's
it's
kind
of
a
slow
going
away
to
get
our
broken
links.
Fixed
and
I.
Don't
have
time
to
do
this,
but
I'm
wondering
if
anybody
does
have
time
to
take
this
as
a
project
to
just
you
know,
run
run
a
link,
checking
tool
across
the
entire
dock
set,
and
just
you
know,
take
the
time
to
try
to
fix
all
the
broken
link.
I
Sense
I
think
it
would
save
us
a
lot
of
time
of
process
thing.
You
know
each
time
we
have
to
process
one
of
these
PRS.
We
have
to
kind
of
understand,
what's
been
done
and
make
sure
it's
been
done
correctly
and
sometimes
a
personal
six,
a
link
at
its
source,
but
there
that
that
bad
link
is
probably
elsewhere.
So
probably
we
want
to
also
update
the
redirects
file
on
that.
We
don't
always
see
that
that
has
been
done.
I
I
I
D
I
D
I
Okay
and
I
think
you
know,
the
idea
of
weekly
or
monthly
is
good,
but
right
now
we
have
a
bunch
of
them
that
are
broken
because
of
the
Hugo
migration.
You
know,
as
best
we
tried
to.
You,
know
to
always
redirect
finally
shuffle
things
around
and
we
still
are
shuffling
I-
think
a
bunch
of
the
ones
we're
seeing
now
are
from
the
migration.
I
A
H
D
D
We're
talking
to
you
see
and
Jennifer's
on
the
line
now.
Is
there
any
update
on
the
reference
box,
because
we
keep
getting
new
issues
about
the
reference
box
and
I?
Think
we
have
a
few
PR
is
open
right
now
and
I
think
some
of
those
things
might
be
conflicting
with
each
other
or
redoing
each
other's
works.
I
I've
been
trying
to
look
at
those
PRS
and
issues
as
I
get
a
chance,
but
I
haven't
haven't
made
any
any
forward
progress
that
there's
one
that
is
in
the
kubernetes
incubator
reference,
Docs
repository
and
I
still
just
don't
understand
what
what
Qingming
is
trying
to
do
there
it.
It
introduces
twenty
thousand
files
into
that
repository
and
it
touches,
and
it
seems
to
be
changing
reference
Docs
that
aren't
even
generated
from
code
that
is
in
that
repository
and
so
I
am
just
baffled
by
it.
J
Follow
up
a
little
bit
there
Steve
did
you
look
at
Jimmy's
last
comments
in
that
er?
He
has
a
long
long
comment
that
he
submitted
over
the
weekend
and
I
just
took
a
quick
look
at
it
TLDR
for
anybody
else
on
the
call
who's
interested.
It's
not
actually
adding
all
those
files
he's
getting
rid
of
enduring
the
trouble
is
the
dip
is
so
big.
You
can't
tell
that
it's
removing
files,
instead
of
adding
them,
there's
a
whole
lot
more
information
in
the
PR.
J
Maybe
you
and
I
can
talk
about
it
separately
later
this
week,
although
this
is
a
PR,
that's
about
the
the
tooling
for
generating
all
of
the
reference.
Docs
and
I
try
to
come
up
with
a
compromise
solution
for
the
direct
for
the
cube,
ATM
docks
anyway
for
the
duration
of
one,
but
eleven
and
I
think
all
I
did
was
make
communication
worse.
J
That
is
that
this
sick
cluster
lifecycle
see
understand
what
I
was
saying:
I'm,
not
sure
that
you
men
did.
He
was
concerned.
You
still
can't
diff,
it's
a
muddled
I
will
see
whether
I
could
do
something
to
help
without
making
a
month
a
meta
metal.
There
but
I
have
not
had
a
chance
to
look
at
any
of
this
stuff
today
and
they
probably
won't
until
tomorrow.
J
J
I
Something
I
truly
don't
understand
is
that
that
repository
is
only
for
two
things:
it's
for
the
kubernetes
api
reference
Docs
and
it's
for
the
coop
control/command
reference
Docs.
It's
not
for
other
things,
like
the
the
single
ref
pages.
We
have
for
kube
api
server,
controller
manager,
scheduler,
and
yet
those
things
seem
to
be
in
the
discussion
in
that
PR.
That's
what
strikes
me
is
just
really
weird
yeah.
J
I
J
A
J
A
J
I
mean
I,
try
to
I
try
to
mention
this.
This
is
complicated
by
the
fact
that-
and
this
might
be
useful
for
everybody-
to
know
to
the
cube,
ATM
docks
and
I'm
partly
responsible
for
this
I
still
think
it
was
a
good
idea,
but
it
they
they're
not
like
any
of
the
other
dogs
record
stocks,
because
they
are
partly
manually
serrated,
you
know
written
by
and
in
work
down,
and
then
those
markdown
files
include
generated
snippets.
J
So
they
include
bits
and
pieces,
it's
what
one
might
otherwise
call
the
committee
and
record
Knox.
This
was
part
of
a
proposal
that
actually
went
true.
Literation
was
last
in
2017
and
was
discussed
by
this
group
and
thus
against
a
cluster
lifecycle,
mostly
fabrizio
panini,
but
it
was
really
the
result
of
collaboration
between
the
two
SIG's
seemed
like
a
great
idea.
J
It
has
not
been
an
issue
until
now
when
sig
cluster
life
cycle
is
doing
a
lot
of
work
on
cue,
medium
and
they
are
updating
both
the
code
and
the
manually
curated
Docs,
and
so
they
I
think
quite
properly
are
submitting
the
generated
Docs
as
part
of
their
pull
request
for
reference,
so
that
you
can
see
with
everything.
That's
changed
only
you
can't
in
the
generated
bits
and
so.
J
Now
we
are
in
this,
like
betwixt
in
between
state
I,
suggested
that
they
not
worry
about
it
just
go
ahead
and
generate
without
the
pretty
tables
in
the
old
ugly
inline
formatting
way,
but
that
still
doesn't
want
if
any
changes-
and
it's
certainly
true-
that
it
can
be
confusing.
Looking
at
the
generated
files
in
a
pull
request,
because
you
start
thinking
to
change
the
generated
file,
I
thought
that
I
was
taking
care
of
all
of
this,
but
of
course
the
PRS
are
wide
open.
J
So
this
was
not
clearly
communicated
and
so
there's
a
couple
of
different
models
here,
I
hope
I'm
explaining
them.
Clearly,
you
know
what
is
whether
generated
Docs
should
ever
show
up
in
a
PR.
You
know
during
the
release
cycle.
I
got
interesting.
Okay
and
the
other
is
this.
This
tooling
issue
which
exists
for
other
generated
Docs
too,
but
has
surfaced
during
the
release
cycle
because
of
the
qadian
I.
A
A
D
If
questions
come
up,
you
can
pay
me
the
next
one
is
probably
for
you:
Zach
I,
think
that
the
kubernetes
organization
is
doing
something
where
they're
adding
security
contracts
for
this
repo,
and
somebody
raised
a
PR
or
at
least
claims
an
issue
or
raised
an
issue
to
do
it
for
Docs,
but
they
needed
to
know
through
those
connections,
see
and
I
suspect.
It
probably
would
be
somebody
at
Ella,
okay,.
A
D
Is
probably
gonna
be
a
quick
one?
I
saw
a
PR
where
somebody
had
made
some
changes
around
an
area
in
a
file
that
included
some
GCE
commands.
They
didn't
add,
GCE
comments
that
I
just
saw
it
when
I
was
reviewing
the
PR,
and
my
gut
feel
is
that
we
shouldn't
have
to
see
commands
in
our
open
source
Docs,
because
we
they're,
probably
kubernetes
Feli,
commands
to
do
the
same
things.
I
So
yes,
I
agree.
We
shouldn't
be
doing
that.
A
lot
of
that
stuff
is
I,
don't
know
if
what
you
were
looking
at
is
old
or
new,
but
there's
a
lot
of
that
kind
of
thing.
That's
just
kicking
around
because
it
was
it's
old
stuff,
so
I.
If
there
were
places
where
we
have
to
mention
cloud
providers,
because
we
need
to
say
you
know,
here's
the
way
you
hook
kubernetes
into
some
hook,
a
cloud
provider
into
kubernetes,
but
I
think
that
anytime
we're
going
to
show
a
a
cloud
specific
command.
D
Maybe
we
if
we
do,
have
to
have
a
command
for
some
reason.
Maybe
we
should
try
to
include
a
command
like
more
than
one
so
that
we
don't
seem
to
be
favoring
one
specific
cloud
provider
like
if
we
had
to
say
how
they
do
it
on
DCE.
Maybe
we
should
say
how
to
do
it.
Also
on
I,
don't
know,
that's
just
one
idea.
If,
if
there
was
a
time
that
we
absolutely
had.
I
A
Cloud
so,
rather
than
point
to
that
kind
of
specific
level,
guidance,
I
would
say
looking
for
the
style
guides
in
terms
of
how
we
would
express
that
in
the
style
guide,
I
I
think
that's
a
good
implementation
of
a
larger
strategy,
and
for
that
larger
strategy,
I
would
say,
make
make
make
Doc's
as
platform
agnostic
as
possible,
link
to
provider,
link
to
provider,
specific
information
and
once
I
think
to
provide
a
specific
information
and
be
as
platform
agnostic
as
possible.
I
could.
D
Imagine
a
case
like
a
path
like
a
walkthrough
in
which
we
choose
a
cloud
provider
to
use
for
that
particular
path,
but
in
that
case,
I
think
it
would
be
great
for
us
to
like
in
the
intro
to
that
say
something
like
for
the
purpose
of
the
task
really
doing
low
cloud.
You
could
also
do
this
have
some
a
different
cloud
provider
so.
A
And
I
would
like
to
make
sure
that
that
partnership,
whenever
we
do
with
with
like
cloud
specific
tutorials,
which
I
think
can
be
a
great
idea
but
to
make
sure
that
the
ownership
or
freshness
of
that
content
goes
to
the
working
group
to
the
cloud
provider
working
group.
Rather
than
than
resting
solely
on
sig
Docs.
A
B
So
I
just
want
to
update
you
so
a
couple
Fridays
ago
we
did
like
sort
of
a
Doc's
print
style
like
bug
bash,
and
we
addressed
about
half
the
issues
and
and
we
were
able
to
close
and
merge
like
the
PRS
that
were
already
open
so
and
then
I
think,
and
then
we
opened
a
bunch
of
new
PRS
to
address.
You
know
some
of
the
issues
so
I
think
we
did
a
pretty
good
job.
B
B
A
A
B
I
B
B
I
K
C
I
One
looks
tough
well,
you
know,
there's
one
example
where
we
discovered
that
if
you
have
a
big
directory
and
in
that
directory
there
are
individual
topics
and
also
sub
directories
yeah.
That
was
it
that
you
can't
you
can't
mix
those
by
using
weights.
You
have
to
have
all
your
individual
topics,
first,
followed
by
all
your
sub
directories
and
you
can
use
weights
to
order
the
list
of
individual
topics.
You
can
just
wait
to
order
the
list
of
sub
directories,
but
you
can't
mix
them
together,
but
I.
Don't
think
that's
a
you
know.
I
B
A
A
K
On
this,
yes
on
this
topics
act,
it
was
actually
one
of
my
bullet
item,
so
I'll
kind
of
move.
My
bullet
item
up
but
I
think
think,
having
Bjorn
do
just
a
little
bit
of
a
code
walk
through
code
review
of
sort
of
the
pieces
of
Hugo
that
he
used.
You
know
it's
got
this
hierarchy
of
templates
and
an
understanding
how
he
did
the
hierarchy
of
templates.
K
You
know
how
he
designed
his
his
base,
HTML
or
whatever
that
the
other
things
inherit
from
getting
more
insights
there
I
think,
would
be
helpful.
I
learned
a
lot
by
doing
a
lot
of
reviews
of
folks
who
are
making
changes,
particularly
you
know,
like
Steve.
He
was
putting
in
a
lot
of
the
waiting,
so
I
learned
a
lot
by
doing
that
and
and
then
just
learning
as
much
Hugo
as
I
could,
but
it
might
help
a
lot
of
these
folks
to
kind
of
jump,
start
being
able
to
say.
K
D
Like
a
cheat
sheet,
there's
a
little
bit
in
the
docs,
but
I
guess
maybe
be
learn
at
is
but
like
I,
don't
really
understand
the
code
short
code
and
like
I,
see
like
as
where
people's
see
ours
change
things
that
they
didn't
need.
A
change
like
where
they're
using
the
wrong
tag
to
end
captures
and
things
like
that,
and
it's
been
kind
of
a
lot
of
detective
work
to
figure
out
in
those
cases.
D
What
the
correct
way
is
look
I've
got
it
now
for
a
lot
of
them
because
of
your
news,
but
like
your
code,
room
I,
don't
necessarily
understand
that.
Well
and
like
we've
found
some
bugs
look
with
lists
and
indication,
and
things
like
that
and
the
order
hasn't
really
weighed
in
even
when
we
finger
monkey
ours,
so
I
think
maybe
it
would
be
good
for
him
to
have
it's
not
like
a
office
hours.
Then
at
least
some
kind
of
a
presentation
where
we
can
ask
some
questions.
A
A
C
B
I
that
so
we
used
to
in
the
net
defied
build
process
specifically
a
like
a
no
index
like
so
that
the
older
versions
of
the
site,
you
know,
like
1.9,
don't
get
indexed
in
the
google
search,
but
I
noticed
that
now
with
hugo
it
we
actually
don't
you
do
with
those
build,
commands
and
nullify
it's
all
controlled
through
the
tamil
file.
But
I
didn't
see
anything
in
that
that
turn.
B
You
know
that
added
the
no
index
thing
so
I
just
want
to
the
idea
I
didn't
have
time
to,
but
we
need
to
check
with
him
if
that's
been
reimplemented,
somehow
or
or
if
we
actually
or
we,
where
we
need
to
re-implement
it,
because
I
don't
want
all
the
various
versions
kind
of
cluttering.
The
google
search
results
andrew
I'm
gonna
move
that
comment.
A
Down
under
under
the
the
larger
bullet
point
for
feedback
from
that,
because
I
agree:
no
indexing
previous
versions
is
really
important
for
user
happiness.
So
that's
yeah,
let's,
let's
definitely
figure
out
how
how
to
implement
that,
whether
it's
implemented
and
make
sure
we
understand
those
particular
mechanics.
K
This
one's
quick
and
easy-
you
know
I
was
doing
bug
wrangling
last
week
with
Steve,
Perry
and,
and
you
know,
obviously
the
ones
I
try
and
grab
at
once.
Were
you
know
somebody's
trying
to
update
through
a
PR
the
blogs,
because
you
know
it's
real
easy
response.
You
just
say
I'm
really,
sorry,
but
who
you
can't
do
that
through
this
review
process
for
not
the
right
reviewers,
you
know,
could
you
could
you
please
go
to
the
blog
and
ideally
do
it
there,
which
I
thought
there
was
a
way
to
provide
feedback?
K
Finally,
I
got
burned.
Somebody
called
me
on
and
said,
there's
no
way
to
provide
feedback
on
that
blog.
You
know,
there's
no
place
to
put
comments,
and
so
I
did
respond.
Okay!
Well,
you
should
go
to
the
authors,
but
I
found
that
really
interesting
was
that
the
best
answer
is
try
and
go
back
to
the
author.
Go
back
to
the
owners
rather
like
there
was
a
list
of
owners
for
the
blog's.
K
A
So
the
mechanics
for
blog
posts,
in
particular,
there's
if
you
look
in
the
the
blog
directory
there
is
a
blog
specific
owners,
file,
yeah
and
right
now.
The
best
way
to
provide
feedback
on
a
particular
blog
post
is
for
is
for
somebody
to
leave
a
comment
on
that
particular
post
and
if
they
can,
but
you
mean,
is
there
a
way
to
leave
a
comment.
A
Mean
on
the
blog
or
well
they
have
to
through
github,
yes,
but
there's
no
right
now,
there's
no
like
from
the
blog
itself
from
like
from
the
website
to
get
home
from
the
blog.
So
it's
we
basically
right
now
we're
telling
people
to
cut
your
own
cut
your
own
hatch
down
into
the
space
underneath
the
stage,
and
you
figure
out
I
figure
out
for
yourself
how
the
works
get
in
there
and
make
a
comment,
and
it
would
be
nice
to
provide
a
clearer
pathway
for
folks
to
leave
blog
feedback.
K
Yeah,
maybe
some
type
of
breadcrumb
at
the
bottom
of
every
blog
article
I,
don't
know
what
it
would
say:
cuz
that
does
sound
kind
of
hard
I
mean
most
people,
I'm
just
typing
a
little
thing
and
hit
post
and
we'll
probably
lose
a
lot
because
it's
involving
it
well
I
couldn't
find
a
way
for
them
to
find
a
place
to
easily
comment
and
get
help.
I
guess
that's
not
too
bad.
I,
don't
know
that.
A
That's
if
you
wanted
to
open
it,
if
you
wanted
to
open
an
issue
about
that
Sarah
Conway,
caitlin
Barnard,
any
of
the
entries
from
the
owners
file
for
the
blog
would
be
great
people
to
mention
in
the
ope.
With
an
issue
saying
hey,
it
would
be
really
awesome
if
we
could
provide
like
on
breadcrumb
trail
burn.