►
From YouTube: GitOpsCon Committee - April 06, 2021
Description
Meeting notes: https://docs.google.com/document/d/1hxifmCdOV5_FbKloDJRWZQHq0ge-trXJKF-BgV4wHVk/edit#heading=h.s5dinyjyr1t
A
Okay,
we
are
now
recording
for
anyone
coming
to
this
later.
Welcome
to
the
april,
6
2021
get
ops
con
cfp
committee
meeting
the
first
one
in
the
series
and
we
we
are
joined
by
the
lovely
people
here.
There's
a
transparency
there
inside
of
the
get
ops
working
group
meetings
on
today's
meeting
notes,
there's
a
link
to
a
github
issue.
A
And
then
so
I
guess
I'll
just
go
ahead
and
since
I
put
my
thing
on
there,
first
I'll
go
ahead.
I
really
just
wanted
to
to
point
folks
to
this
committee
meeting
committee
tracking
issue
that
christian
made
it's
fairly
straightforward.
It's
just
making
sure
that
there's
some
transparency
as
to
you
know
who's
doing
what
and
how
that's
how
that's
working.
A
A
A
And
then
I'll
reach
out
I'll
reach
out
to
tama
great.
I
didn't
really
have
anything
to
speak
to
really
about
this,
except
for
just
trying
to
make
sure
that
you
know
we
do
have.
We
do
have
that
project
board
for
for
helping
us
to
organize
things.
I
hope
it's
useful
for
people
if,
if
you
think
we
should
change
it
up
totally
change
it
up
in
any
way,
but
that's
just
housekeeping
stuff.
So
I
guess
I'll
just
wait
until
anyone
has
ideas
on
their
own
about
that
and
yeah.
B
Well
and
more
house
a
little
bit
more
housekeeping,
I
mean
scott
figured
out
that
if
you're
gonna
use
a
project
board,
it's
actually
easier
to
create
an
issue
and
put
use
that
as
the
card,
because
we
were
doing
it
another
way
before
it's
fine,
if
you're
just
putting
in
cars
just
like
to
jot
down
your
ideas.
But
eventually
you
know,
we
found
it
easier
to
retroactively,
go
and
create
an
issue
and
then
make
that
the
actual
thing
that
you're
tracking
in
the
project
so.
D
A
A
A
In
short,
in
short,
for
those
that
don't
have
accounts,
it's
the
most
important
thing
to
keep
in
mind.
Is
that
the
the?
Why
we're?
Even
why
even
bother
I
mean
I
could
just
pull
up
the
xkcd
chat
systems
comic,
and
we
can
all
have
a
laugh
at
like
you
know
how
many
different
systems
we
are
thinking
about
using.
The
only
reason
I
suggest
using
key
bases
is
for
is
not
to
replace
everything
that
we
have.
We
have
a
slack
channel.
We
have.
We
have
github,
we
have
projects
for
things
like
that.
A
It's
really
for
right.
Now.
It's
really
only
in
case
there's
sense.
Any
information,
that's
sensitive,
that's
it
so
before
we
have
a
an
open
source,
one
password
account
which
we'll
get
pretty
soon,
there's
a
technique
to
to
to
encrypt
files
there
and
do
file
sharing,
and
it
can
be
on
a
team
basis.
So
I
have
a
I'll
post
it
again,
but
I
have
a
a
good
github
gist.
That
shows
how
to
do
that.
A
I
set
it
up
for
the
helm
team
and
that's
what
we've
been
using
for
for
certain
types
of
files.
You
know
yeah,
the
only
other
thing
is
to
know
if
you,
if
you
the
main
the
main
value
that
key
base
brings,
is
it
helps
to
reduce
the
possibility
that
someone
can
get
a
hold
of
something
important,
so
its
main
value
proposition
is
is
being
able
to
know
that
you
can
trust
a
person
to
have
greater
and
greater
levels
of
trust
to
know
that
a
person
is
actually
who
they
say.
C
A
Yeah,
this
is
just
info
for
anyone
either
listening
in
or
like
on.
The
call
who
doesn't
have
an
account
yet,
and
the
reason
I'm
mentioning
that
is
is
that,
if
you
are
do
need
to
make
an
account
still
it's
best
not
to
put
an
organization
name
or
anything
like
that
in
there
key
base,
primarily
not
just
rounds
upon,
but
really
its
whole
value
proposition
is
that
you
have
one
account
that
you
prove
you
are
who
you
are
so
just
make
it
you
and
and
yeah.
C
A
You
mean
google,
google
drive
or
whatever
google
google
drive
the
only
the
reason
I
suggest
against.
That
is
it's
it's
it's
known
to
cause
problems
specifically
in
cases
where
you're
doing
demos
yeah
I
mean
no
one
else
should
be
able
to
get
in
there
without
google's
im,
but
it
comes
right
up
in
search
you're
doing
a
demo
and
all
of
a
sudden,
like
your
keys
for
this
like
whole
shared
team
show
up
and
and
now
everybody
has
to
rotate
our
keys.
A
A
A
C
A
A
B
Yeah,
so
thanks
scott
yeah,
I
for
that
for
those
housekeeping
things
here,
so
the
the
I
actually,
I
leaned
on
scott
to
try
to
like
put
all
this
together,
because
the
cfp
is
closing
soon
right
next
next
next
friday,
and
then
it
actually,
we
don't
have
that
much
time
to
make
decisions,
because
you
know
there's
only
two
weeks
until
the
event
after
it
closes,
so
we
actually,
you
know
we
need
to
start
like
doing
some
of
this
stuff.
We
actually
need
to
do
it
like
in
a
few
days
right,
yeah.
B
To
get
I
wanted
to
get
like
this
meeting
this
weekend
next
week
to
go
over
the
the
criteria,
so
I
think
leo
put
together
a
good
good
draft
by
the
way,
thanks
leo
for
for
doing
that,
because
this
is
my
first
time
doing
this
here
and
so
it's
linked
in
the
the
agenda
doc.
But
let
me
also
link
it
here
in
the
chat
for
everyone
else
to
see.
B
C
C
F
D
C
D
Yeah,
so
this
is,
this
is
basically
a
kind
of
a
distilled
version,
much
more
to
the
point
based
off
of
the
cncf
guidelines
and
basically
there's
like
a
handful
of
concepts
that
are
relevant.
I
think
the
fact
that
there
has
to
be
some
level
of
control
in
terms
of
conflicts
of
interest,
so
it's
it's
very
broad
in
in
the
context
of
just
like
if
you
work
with
the
person
that
submitted.
D
Then
the
other
one
was
the
fact
that
this
grading
mechanism
of
one
to
five
porter
excellent,
that
I
just
copied
over
and
the
concept
of
categories
right,
that
there
will
be
multiple
categories
that
would
be
evaluated.
So
I
those
are
not
verbatim.
What's
in
the
cncf
they're
kind
of
like
a
little
bit
of
an
interpretation
of
my
own
and
and
what
I
think
we
should
go
for,
and
then
I
added
something
that
was
not
in
the
cncf
exactly
and
it's
kind
of
like
the
mechanism
to
affect
to
effectively
average
all
those
different
scores.
D
Because
what
I
have
in
mind
is
we
evaluate
each
one
independently?
We
assign
a
score
for
each
one
of
those
categories
and
we
basically
average
those
four
scores
into
a
kind
of
global
cfp
value.
We
basically
order
them
descendingly
on
on
global
value
and
evaluate
on
on
broader
criteria.
Right
like
is
there
a
lot
of
redundancy
in
subjects
across
the
top
selections
right?
F
C
Actually,
I
disagree,
I
think,
for
stuff.
That's
gonna
show
on
kubecon
like
accepted
cucumber
talks.
I
would
limit
more
at
most
to
a
slot
in
the
lightning
talks.
There's
a
kind
of
teaser
preview,
summary
of
the
long
talk
in
kubecon,
because
everybody
who's
in
our
zero
day
will
be
also
at
kubecon,
and
I
think
we
should
give
a
chance
to
the
people
who
have
to
say
something
very
much
about
get
ops,
which
is
maybe
also
not
kubernetes,
related
and
also
a
chance
to
people
who
didn't
get
into
kubecon
with
their
talk.
F
I
think
that's
reasonable
right,
like
rack
and
stack
them
priority
wise
if
they're
at
kubecon
right
like
bottom
of
the
list
kind
of
deal,
I
think
that's
reasonable,
but
I
would
not
like
just
say
flat
out
like
oh,
you
got
accepted
kubecon.
No,
we're
not
going
to
have
you
because
it's
very
easy
to
do
the
same
talk
two
different
ways.
In
my
opinion:
okay,.
A
B
A
professional
speaker
right,
I
I
I
tend
to
agree
with
with
both
those
points,
so
I
I
think
they
should
have
loyal,
lower
priority,
but
not
flat
out
say
no,
that's
you
know.
That's
that's
that's
my
my
first
instinct
is
to
be
like.
Okay,
like
we
should
give
original
talks
and
things
specific
to
get
ups,
for
you
know
the
ones
that
we
like
first
right
and
then
like.
D
I
agree
just
just
just
to
be
clear.
I
don't
think
kind
of
what
I
was
implying
is
not
saying
no
to
any
talk
that
has
overlap,
but
that
it
has
to
be,
or
that
I
would
suggest
it
is
a
component
that
is
incorporated
in
the
decision
process
right.
It's
that
we
should
be
mindful
of
what
is
going
to
be
presented
afterwards,
not
just
what
is
going
to
be
presented
in
the
zero
day.
To
some
degree.
I
guess
so.
D
F
D
The
stop
for
beginners,
there's
kind
of
entry-level
stuff,
more
advanced
things,
there's
ideally
a
balanced
variety,
and
that
there
are
no
major
overlaps
or
duplication
with
talks
that
are
going
to
be
presented
in
kubecon
and
the
kubecon
event.
I
don't
know
if
that's
a
a
faithful
way
to
word
what
we're
talking
about.
So
I'm
all
open
for.
A
F
F
B
C
A
B
F
B
F
B
A
A
D
Know-
and
the
reason
I
pointed
out
three
members
is
because
I
think
it's
it's
a
it's
a
good
starting
point.
In
other
words,
I
think
anything
below
three
members
is
not
fair.
Of
course,
a
single
review
doesn't
doesn't
go
two.
There
is
no
no
kind
of
breaking
in
consensus.
You
know
what
I'm
saying
two,
I
think
three
is
likely
the
minimum
number
to
get
a
fair,
just
average
across
multiple
perspectives.
D
E
So
yeah,
I
wanted
to
make
a
couple
of
suggestions,
so
I
I've
learned
a
lot.
I've
been
on
the
programming
committee
for
the
devops
enterprise
summit,
where
we
get
hundreds
of
submissions
and
have
a
programming
committee
about
the
same
as
the
size
of
this
one
and
so
a
couple
of
rules.
I
think
that
are
super
helpful
is
first
of
all.
I
don't
have
any
objection
to
there
being
more
than
three
reviews
and
I
think
that
that's
fair
we're
gonna
average.
So
it's
not
total
score.
E
It's
we
have
to
average
across
that,
so
we
can.
We
can
make
up
for
the
difference
in
three
versus
five
or
something
like
that.
So
so
I
wouldn't
want
to
constrain
that.
The
second
thing
is
that
we
have
a
similar
rule,
a
slight
slight
variation
where
we
say
any
talk
that
ends
up
with
an
average
of
in
our
averages,
the
system
that
we
use.
E
The
last
comment
that
I'm
going
to
make,
though,
is
that
I
had
I
had
been
on
the
programming
committee
for
some
w3c
conferences,
which
were
blind
you,
the
system
was
set
up
so
that
you
couldn't
see
who
was
presenting.
You
couldn't
see
you,
you
were
based
on
the
proposal
alone
and
the
first
year
that
I
did
get
devops
enterprise
summit.
I
did
that.
I
didn't
look
at
who
the
submitters
were
and
things
like
that,
and
then
I
realized
over
the
last
several
years
that
there's
a
lot
of
value
in
that
it's
like.
E
E
E
We
are
responsible
for
creating
a
fantastic
program,
and
so
it's
okay
for
me,
for
example,
if
we
end
up
selecting
something
that
ends
up
with
an
average
score
of
three
and
a
half
over
something
that
is
an
average
score
of
four
and
a
half
because
there's
some
criteria
that
will
make
it
a
better
event.
We
already
talked
about
one
of
them,
this
talk's
already
on
the
kubecon
agenda.
So
let's
make
room
for
this
other
talk,
which
is
really
going
to
enhance
our
schedule
and
enhance.
You
know
the
program
that
we
put
together.
E
F
C
A
C
Now
I
can
share
okay,
so
what
I
did
is
I,
while
we
were
all
talking,
wait,
I
went
ahead
and
I
set
up
a
protection
on
the
raw
data
so
that
nobody
would
accidentally
change
it.
If
you
like
go
ahead
and
change
it,
there
comes
a
warning
that
says:
don't
do
that
nice
and
gives
you
a
chance
to
bail
out,
and
I
created
a
sheet
called
review
which,
in
this
gray
area,
which
I
actually
forgot,
I
have
to
protect
it
as
well.
C
Just
a
second
we'll
have
the
same
warning
so
in
the
gray
area
on
the
left,
we
have
just
the
raw
data
mirror,
but
without
personal
information,
so
that
it's
the
actual
stuff
that
we
should
be
reviewing
initially
and
I
created
a
little
table
where
you
can
now
put
some
grades
here.
Like
I
don't
know,
four
one:
five
yeah
like
four
four
four.
I
think
you
get
the
idea-
and
I
just
realized
that
here
we
see
a
bit
too
much.
C
B
C
C
Here's,
a
total
which
I
was
just
about
to
do,
here's
a
column
for
cubecond,
repeat-
and
my
thought
was
that
here
in
the
last
column,
we
can
put
some
reasoning,
thoughts
or
whatever
notes
we
want
to
keep
for
ourselves
about
this
talk
I
don't
know
special
minority
speaker.
We
should
take
it
nevertheless,
or
I
know
him,
he
actually
doesn't
speak
english.
What
should
we
do?
I
don't
know
yeah
whatever
notes
we
should
take
into
consideration
and
then
we
can
add
further
columns
on
demand.
B
E
C
E
A
B
E
When
people
can
use
the
comment,
they
can
stick
a
comment
into
the
cell
if
they
want
to
say
you
know,
I
think
this
is
particularly
relevant
or
this
is
totally
irrelevant,
or
something
like
that.
You
know
some
comments
that
even
I
find
that
making
the
comments
reminds
me
when
I
go
back
like
when
I
thought
through
this,
what
why
did
I
score
it
this
way?
So
I
find
that
helpful.
C
E
E
We
can
do
it
just
in
time,
but
while
you're
editing
it,
I
think
a
column
for
accept
reject
is
helpful,
and
one
of
the
other
things
that
I
want
to
circle
back
on
is
that
when
we,
I
think
that,
having
interaction
with
the
submitter
to
help
them
refine.
So
one
of
the
things
that
we
do
fairly
frequently
is
when
we
find
that
there's
two
people
from
the
same
company,
for
example,
that
have
submitted
talks
and
they're
closely
enough
related.
E
We
might
circle
back
with
them
and
say
how
how
about
you
do
a
joint
talk
or
they
might
be
across
two
different
companies,
and
you
know
how
about,
if
you
bring
a
business
leader
with
you
to
this,
I
don't
think
that's
relevant
here,
but
the
this
is
the
type
of
thing
that
we
do
as
well.
So
as
we
go
through
and
evaluate
these
things,
I
think
it's
fair
for
us
to
to
circle
back
and
request.
More
info,
like
I
wouldn't
say,
take
a
super
thin
abstract
and
do
a
lot
of
work
to
figure
out.
E
C
And
then
kind
of,
but
also
give
them
a
very
short
deadline.
Let's
say
from
friday
till
sunday
or
whatever
to
please
submit
that
information,
and
then
we
can
ask,
for
example,
about
if
it
has
been
accepted,
a
kubecon
if
it's
a
repeat
talk
that
they
gave
only
20
times
stuff
like
that
and
also
like.
What's
how
relevant
are
the
githubs
principles
to
their
talk?
F
E
C
Yeah
sugar
coated
no!
Well,
I
mean,
I
think,
everybody
I
agree
with
and
I
don't
think
we
need
to
have
a
conference
that
will
show
more
ci
of
stocks
unless
it's
a
talk
that
spends
two
minutes
on
explaining
why
their
old
ci,
ops
setup
wasn't
good
enough
and
then
explaining
how
they
rebuild
that
into
get
ups.
Sorry,
I'm
hogging
the
meeting.
A
You're
fine,
so
I
just
raised
my
hand
for
this
quick,
just
quick,
two
cents
is
that,
as
we've
said
before,
I
wanna
I
I
personally
want
to
be
cautious
about,
I
mean
look.
We
have
a
process
set
up
to
help
with
to
help
average
our
biases
right
and
help
keep
each
other
accountable
on
things
that
shouldn't
eat.
D
Yeah,
I
just
want
to
kind
of
like
echo
what
scott
said,
because
I
think
part
of
the
whole
exercise
here
is
to
also
understand
exactly
how
what
get
ups
means
in
the
wild
you
know
and
and
how
it's
being
applied
and
how
it's
being
understood,
because
it's
still
in
flux,
it's
still
being
defined
and
it
should
be
defined
in
the
way
that
incorporates
what
what
the
real
use
cases
out
there
really
bring
forth.
So
I
think
I
think
I
just
want
to
kind
of
resonate
to
what
scott
said.
D
D
A
A
B
Yeah
this
is
this
is
very
useful.
I
don't,
I
think,
a
lot
of
this.
I
think
we've
actually
talked
a
lot
about
about
this.
I
don't
know
if
we
want
to
continue
on
this
track
or
or
wait
or
wait
for
it
for
the
next
sync
up.
I
do
want
to
talk
about
the
timeline
that
we
have
for
voting,
because
it's
a
very
tight
timeline
and
the
the
the
timeline
is.
Essentially,
we
close
the
cfp
there's
two
weeks
and
then
we
have
the
event.
B
B
F
B
D
F
D
Or
or
we
could
help
them
out
because,
for
instance,
I
did
a
talk
that
the
way
that
they
recorded
was
they
actually
just
brought
me
into
a
zoom
call
and
they
recorded
the
zoom
call.
So
if,
if
we
do
ad,
we
could
kind
of
like
help
them
out
and
avoid
the
kind
of
round
trip
latency
of
having
to
request
a
redo
and
and
just
say,
hey,
you
know
what
your
license
is:
no
good.
You
kind
of
go
ahead
like
this
and
not
waste
time.
If
we
wanted
to
coordinate
recordings
with
them,
that.
H
D
F
D
D
E
E
I
think
that's
important
for
us
to
do
today
so
that
we
can
all
plan
and
make
sure
that
our
calendars
are
open
enough,
so
that
we
can
dedicate
the
time
to
do
the
the
submissions
we
can
start
reviewing
them
as
as
they
come
in.
But
we
should
understand
that
timeline
and
then
I
also
think
it's
super
important
that
today
that
we,
at
least
at
a
high
level,
talk
about
the
slot
count.
E
C
B
E
F
D
That
one,
so
if
you
go
all
the
way
up
all
the
way
up,
so
I
got
a
cfp
open
and
close
and
then
kind
of
basically
took
it
backwards
and
try
to
allow
it
enough
time
for
dry,
runs
and
testing
how
long
we
have
to
get
the
recordings
blah
blah.
So
that's
kind
of
like
just
an
estimate
that
I
would
like
us
to
kind
of
walk
through
meant.
E
E
B
Yep
yeah,
I
think
the
deadline
for
the
21st-
I
think
that's
that's
fine.
I
think
I
think
the
21st
is
when
we
should.
I
think
I
want
to
push
it
back
a
little
bit
so
from
from
your
agenda
here.
Leo
is
actually
like
voting
on
the
1920
and
like
the
21st.
Those
emails
go
out
right
because
the
sooner
they
know
about
it,
the
I
think
the
better,
but
I
don't
know
if
that
gives
us
enough
time
to
like
decide.
B
E
Now
I
believe
that
we
as
a
committee,
I
believe
we
as
a
committee,
should
see
if
we
can,
we
should
calendar
some
time
to
get
together
like
we
are
now
to
discuss.
You
know,
we've
concluded
our
voting,
you
know
and
then
we
need
to
come
together.
I
know
we'll
try
to
do
as
much
asynchronously,
but
there's
it
will
be
so
much
more
efficient
if
we
can
at
least
have
a
couple
of
meetings,
so
maybe
that,
following
monday
tuesday,
we
put
two,
you
know
two
one-hour
blocks
on
the
calendar
for
monday,
tuesday.
F
B
B
B
F
D
F
D
C
F
C
C
E
Sorry,
I
didn't
see
that
yeah
no
problem,
so
since
we
only
have
five
minutes
left,
I
did
want
to
jump
all
the
way
down
to
my
last
thing
on
the
agenda
and
that
is
invited
talks.
Question
mark.
I
have
one
in
particular
in
mind
and
I
would
love
to
hear
if
there's
any
objections,
I
would
like
to
invite
a
talk
from
the
cdf,
probably
from
tracy
miranda
directly,
because
obviously
go
ahead.
Slomo
the.
E
Oh
continuous
delivery
foundation,
continuous
delivery
foundation-
I
will
spell
it
out,
and
and
so
I
would
like
to
invite
tracy
to
be
an
invited,
talk
to
do
a
to
to
really
just
bring
the
communities
together
and
to
show
that
really
to
have
her
talk
about
what
git
ops
means
in
the
context
of
the
cdf
and
and
her
intention,
you
know
her
her
desire
to
have
our
communities
work
together,
so
I
would
love
to
have
that
as
a
15
minute.
Okay,
I'm
seeing
lots
of
plus
ones
awesome
yeah.
Thank
you.
Definitely.
E
D
B
B
D
D
A
C
F
D
F
F
A
B
A
A
Yeah,
I
do
think
that
we
should
I'm
glad
you
mentioned
that
I
do
think.
We
should
be
clear
that
that
this
is
to
be
that
this
is
to
be
made,
or
this
is
to
be
kept.
You
know
between
the
program
committee,
if
there's
any
reason
to
address
it
afterwards
with
someone,
we
can
always
do
that,
but
yeah
that
seems
to
be
standard
practice.