►
From YouTube: CI/CD Group Conversation (Public Livestream)
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
B
A
A
So
it's
not
to
say
that
we
will
never
do
this
or
even
that
we
won't
do
it
soon,
especially
with
a
team
focused
on
an
onboarding.
In
that
experience,
we've
got
both
the
growth
team
and
within
CI
CD,
we're
focusing
on
onboarding
and
making
that
smooth.
But
that's
the
reason,
probably
why
that
we
haven't
prioritized
it
compared
to
some
other
things
in
the
past.
So
it's
actually.
It
was
actually
nice
to
see
it
again,
and
you
reminded
me
of
it
and
I
knew
that
it's
out
there
so
yeah.
A
A
There
was
two
ways
that
I
could
interpret
this
question
and
said:
it's
it
is
away.
Is
why
he
has
to
read
it?
One
is:
when
are
we
delivering
a
new
package
format
and
we
just
delivered
nougat,
which
is
a
primarily
windows-based
package
manager
that
we
just
launched
in
our
12.8
release,
which
was
out
just
a
few
days
ago?
The
other
way
that
I
could
interpret
this
question.
That
I
think
is.
A
Yet
if
you
follow
this
page,
if
you
submit
an
M
R
that
implements
these
five
or
six
things
that
we've
documented
very
clearly,
then
you
know
you
can
contribute
that
and
it
will
work
great
and
it
will
just
be
as
integrated
as
nicely
as
all
the
other
package
manages
and
I
want
to
pause
here
and
ask
if
we
have
Tim
on
the
call
who
can
give
more
details
about
where
we're
at
I
can't
remember.
If
he
said
he
was
a.
C
Well,
I
I
should
just
rewind
and
say
we're
also
working
right
now
on
PI
pi,
PI
integration,
that's
sort
of
the
next
format,
that's
upcoming,
that's
been
requested
by
a
lot
of
our
customers
and
users.
It's
a
fairly
popular
issue
and
we're
planning
on
delivering
we're
doing
an
investigation
right
now
and
then
planning
on
delivering
support
in
twelve
ten
and
twelve
eleven
over
the
next
two
milestones.
C
For
that,
but
separately
on
sort
of
a
driving,
more
contributions,
we've
been
having
community
office
hours,
we've
got
two
contributions
in
progress
now
for
adding
support
and
we've
been
continuing
to
improve
the
how
to
contribute
development
documentation
and
the
the
next
piece
of
that
is.
How
do
we
abstract
away
some
of
that
and
make
a
more
just
a
quality,
Universal
package
manager'
and
we're
starting
to
move
forward
with
that
vision
in
terms
of
setting
up
riverboat
in
virtual
repositories?
A
D
A
So
the
last
thing
I
heard
this
was
still
the
case,
but
what
that
means
is
that
using
the
data
for
smell,
we
are
definitely
over
counting
the
number
of
people,
because
there's
kind
of
duplicate
data
sources-
there's
not
duplicate,
but
there's
disaggregation
of
instances
and
we're
kind
of
estimating
users
based
on
kind
of
abstract
things.
And
so
that's
what
we
mean
by
overestimated
smell.
It
means
that
the
number
is
larger
than
the
actual
number,
but
it
should
trend
towards
the
actual
values.
A
D
Thank
you,
Jason
I'll,
add
just
one
data
point
that
I
know
about
why
we
decided
to
call
it
overestimate.
It
was
that,
typically,
your
usage
ping
data
is
at
the
instance
level.
So
it'll
say
you
know
there
were
this
many
interactions
with
this
part
of
the
product,
but
we
don't
know
that
at
the
user
level,
so
we
just
multiply
the
number
of
users,
that
instance
by
the
fact
that
the
instance
had
the
activity
happening
so
automatically.
Not
every
user
did
that
yeah.
A
D
Great
presentation,
thank
you.
The
slide
9
refers
to
moving
continuous
delivery
to
lovable,
which
I
think
is
the
first
time
we've
ever
moved
an
existing
category
that
wasn't
lovable
to
lovable
super
cool,
to
see
that
I.
Just
wonder
if
you
could
talk
through
like
how
we
came
up
with
the
plan
to
make
it
lovable
and
maybe
what
that
entails.
In
your
mind,
yeah.
A
So
I
think
continuous
integration
is
loveable
already
and
perhaps
pages,
and
not
a
hundred
percent
sure
on
that,
but
yeah
so
I
linked
below
to
where
the
maturity
plan
is
which
has
collected
the
issues
that
that
we
think
are
that
our
inventory
thinks
are
the
ones
that
are
the
most
necessary
for
that
we're
sort
of
iterating.
On
this
this
process
we
a
while
ago,
we
didn't
used
to
even
have
the
you
know.
What
is
the
plan
section?
A
We
now
have
the
list
of
issues,
but
we
don't
really
have
any
narrative
that
ties
them
together,
at
least
that
I'm,
aware
of
or
eat
may
have
one
in
an
epoch
or
something,
but
it
either
way
it
should
be
linked
there.
So
I
can
speak
in
a
general
sense
on
my
own
kind
of
narrative,
Lee
about
where
I
see
that
category
moving
to
and
from
but
yeah
it
isn't
documented.
Currently,
as
far
as
best
I
know
and
like.
D
A
A
We
also
haven't
so
I
know
that
there's
an
effort
to
you
base
maturity
on
some
new
metrics
like
NPS
or
something
else,
I,
don't
know
if
we've
made
a
decision
there
or,
if
that's
still
being
discussed,
but
we
haven't
pinned
that
category
yet
to
a
new
kind
of
like
objective
analyst
or
other
measure
panel.
For
us,
though,
like
just
to
just
kind
of
speak
to
the
category
and
where
we're
headed
with
it.
A
What
we
want
to
do
with
the
CD
category
overall
is,
is
gonna,
be
a
lot
of
building
I,
think
on
Auto
DevOps
and
adding
non
kubernetes
stuff
to
Auto
DevOps.
So
ohrid
has
a
I:
don't
have
you
called
PSE
or
just
kind
of
an
iteration
that
we're
working
on
right
now,
where
she's
looking
to
add
a
composable
template
for
ECS
that
if
you
have
ECS
configuration
in
your
secret
variables,
if
you
have
again
I
can't
memorize,
it's
called
the
manifest
or
whatever
that
the
AWS
expects.
A
When
you're
deploying
to
ECS
that
we
will
auto
detect
the
presence
of
those
things
and
then
just
like
we
do
for
communities,
we
will
run
a
whole
different
set
of
deploy
tasks
and
implement
incremental
rollout
or
whatever
else
that
we
can.
So
as
we
get
closer
to
that
vision
for
CD,
then
we
have
started
to
to
really
achieve
something.
That's
at
the
next
level
and
I
think
is
really
lovable
and
market
leading.
The
other
aspect
that
plays
into
this
is
the
overall
concept
of
progressive
delivery
and
where
progressive
delivery
is
a
broad
concept.
A
So
review
apps
feature
flags
and
just
kind
of
super
super
fine,
grained
control,
integrator
monitoring
and
just
slowly,
slowly
staging
things
out
in
a
very
safe
way
that
you
don't
impact
anybody
that
you
don't
expect
and
wiring
that
all
together,
it
touches
quite
a
number
of
different
categories,
many
of
them
in
release,
but
some
of
them
in
your
area
related
to
monitoring
and,
of
course,
Auto
DevOps
as
well
and
wiring.
All
that
together
is
kind
of
the
the
place
where
the
rubber
hits
the
road
is
in
the
continuous
delivery
category.
E
E
C
E
A
So
great
to
hear
we're
really
trying
to
make
a
focus
on
dog
fooding
across
the
board
and
make
sure
that
we're
auditing
our
features
to
see
who's
using
them
and
and
that
we're
using
them
internally
and
that
we're
not
blocked
anywhere.
So
that's
really
great
to
hear
and
we'll
look
to
replicate
that
approach
if
it's
worked
so
well
within
the
testing
group
yeah.
Thank
you.
James
I'll
also
say
the
same
thing
Brandon
over
to
you.
F
Morning,
I'm
didn't
quite
type
as
fast
as
I
was
gonna.
Ask
the
question
so
I'll
finish
typing
after
ask
the
question
so
love
what
we're
doing
in
CD,
I,
guess
kind
of
the
question
a
little
bit
is
when
we
look
at
visualization
when
I
look
at
the
visualization
and
the
way
that
we
surface
it,
it's
surfaced
very
much
in
stream,
which
obviously
makes
sense.
That's
what
our
value
proposition
is
is.
Okay,
it's
the
next
step
and
CD
in
some
ways
is
like
hey,
we're
rolling
it
out,
but
almost
made.
F
You
know
we
make
it
as
simple
as
possible,
which
generally
I
totally
agree
with.
Wondering,
though,
can
if
you
flip
it
to
a
CD
strategy,
there's
a
whole
bunch
of
operations,
teams
that
we'll
be
looking
at
and
they
kind
of
they
won't
touch
the
CI,
but
once
it
goes
through
yeah
and
hits
that
CD
stage
they're
gonna
want
to
see
all
the
progress
or
you
know
detailed,
progressive
rollouts
and
it
feels
like
we
have
the
right
features,
but
maybe
we're
not
surfacing
them
in
a
way
that
that
user
is
used
to
or
would
expect
it.
F
A
How
y'all,
yeah
and
I
can
give
an
example,
and
you
can
let
me
know
if
I
understood
it.
So
one
of
the
the
things
that
we're
looking
at
doing
next
on
feature
flags
is
having
that
feature:
flag
status,
information
to
be
tied
in
to
the
issues
and
then
in
the
environments
as
well,
so
that
there
would
be
a
view
if
you're,
the
kind
of
person
who
is
maybe
a
tester
and
are
interested
in
like
what's
deployed
to
my
environment
right
now.
And
what
are
the
configuration
settings
if
you
go
to
the
environment?
A
You'll
see
an
overview
there
of
all
the
feature
flags
that
environment
has
and
what
the
current
state
of
them
is.
And
then,
if
you
go
to
an
issue,
if
you're
a
developer,
working
on
an
issue,
there'll
be
something
on
the
sidebar.
That's
like
you
know
here
are
the
feature
flags
that
we
introduced
as
part
of
this
particular
issue
and
here's
a
list
of
all
the
environments
and
what
what
they're
turned
on
or
off
in
those
places.
A
So
we
are
headed
there
and
we're
doing
the
exact
same
thing
with
releases,
and
you
know
making
like
what
releases
in
this
environment.
What
release
is
related
to
this
issue?
You
know
what
milestones
are
related
to
this
release
and
just
wiring
everything
together.
It's
all
a
bit
nascent
right
now
releases
is
probably
the
farthest
along
and
actually
delivering
some
of
that
stuff.
Where
you
can
like
from
a
release,
you
can
go
look
at
a
milestone
that
was
part
of
it
and
see
the
burndown
chart
that
was
related
to
it.
A
Another
really
cool
feature
that
we're
we're
pushing
forward
in
this
direction
is
in
this
release.
We
will
be
able
to
auto
generate
a
change
log
from
the
you
know,
from
MRA's
or
from
the
commits
that
can
be
attached
automatically
to
the
to
the
release
itself
and
that
kind
of
will
come.
You
know
from
the
application
at
large
and
maybe
hopefully
we're
looking
into
dogfooding
that
so
that
part
of
the
release
post
process
that
we
have
could
potentially
be
auto-generated
in
that
same
way
and
then
we're
using
our
own
feature
we're
improving
it.
A
F
I
think
those
are
kind
of
exactly
the
that's.
Some
really
good
examples,
and
thank
you
for
for
highlighting
those
those
are
there
and
I
will
keep
I'll
keep
working
on
finding
users
that
have
that.
As
a
as
a
use
case,
you.
F
Then,
what
we'll
work
to
find
be
able
to
demos
with
them
and
stuff
just
to
get
that
feedback,
but
yes,
I,
think
I
think
that's
a
whole
new
group
of
users.
It
looks
at
the
release
process
again
and
of
just
a
very
different
way
than
than
we
do.
So.
That's
that's
great
to
see
really
encouraged,
be
happy
to
see
those
and
I'll
help
any
way
I
can
on
the
dogfooding
and
finding
me
yeah.