►
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
Alright
looks
like
it's
time
to
start.
My
name
is
Jason.
Lenny.
Welcome
to
the
group.
Conversation
for
a
CI
CD
looks
like
folks
are
starting
to
ask
to
add
some
questions,
but
I
wanted
to
kick
things
off
by
touching
on
the
topic,
slides
and
just
briefly
talking
about
those
I'll
share.
My
screen
here.
A
And
if
we
jump
into
the
topics,
then
one
of
the
big
things
that
we're
doing
right
now
is
planning
for
FY
21
and
how
we
want
to
grow
the
team,
some
of
the
things
that
we're
thinking
about
for
C
ICD
well,
all
of
its
really
oriented
around
the
CIC
D
strategy
that
we
just
recently
published,
and
hopefully
everybody's
had
a
chance
to
look
at.
We
are
going
to
be
doing
a
kind
of
an
internal
session
focused
on
the
design,
engineering
and
other
stable
counterpart
teams
that
work
with
CID
C
ICD.
A
So
if
you're
interested
in
attending
that,
let
me
know-
and
I
can
probably
add
you
to
that-
invite.
But
it's
all
Senator
on
that.
But
some
of
the
themes
that
we're
looking
at
are
focusing
on
onboarding
and
potentially
building
out
a
team
that
can
look
at
onboarding
and
getting
up
to
speed
with
with
git
lab
across
verify,
packaged
and
released.
So
thinking
about
the
whole
CI
CD
experience.
Of
course
we
want
to
grow
the
runner
team.
A
We
recently
split
that
up
from
C
ICD
and
we
want
to
go
it
up
until
a
full,
full
sized
team.
We
also
are
looking
at
how
we
can
support
the
mobile
use
case
better
and
potentially
building
on
a
team
that
are
sorry
group
that
can
focus
on
that
use
case
again,
maybe
crossing
some
of
the
lines
between
the
individual
teams
and
maybe
for
the
mobile
use
case,
even
supporting
things
like
syntax,
highlighting
associated
with
mobile
languages
in
the
web
IDE,
and
things
like
that.
A
Availability
and
features
for
the
container
registry
is
also
a
focus
and
then,
with
all
the
splitting
that
we've
done,
we
have
a
bunch
of
teams
now
that
are
sort
of
partial
size,
and
so
we
want
to
grow
all
of
those
teams
back
up
to
full.
These
are
just
the
ideas
that
we're
discussing
right
now
and
we
still
have
time
to
figure
out
what
exactly
we're
going
to
do.
Your
feedback
is
welcome,
so
so,
if
you
have
any
thoughts,
Windows
and
Mac
renner
work
is
also
progressing.
A
You
can
see
the
links
to
the
epics
here
and
the
issue
where
we're
discussing
pricing
and
how
the
beta
rollout
will
happen
at
the
moment.
All
of
that
is
plans
worth
finishing
everything
for
12.5.
So
again,
super
super
exciting.
Also
recently,
we
had
the
cloud
native
CI
Wave,
where
we
performed
really
really
well.
We
in
terms
of
the
strategy
and
vision
and
product
roadmap.
A
We
scored
a
perfect
5
out
of
5,
and
then
you
can
see
if
we're
up
here
on
the
top
right
doing
well
with
the
individual
cloud
providers
like
Google,
Microsoft
and
AWS
in
there
and
their
solutions.
We're
really
interesting
placed
here
because
we're
the
only
really
independent
operator.
That's
that's!
In
that
space
so
exciting.
A
Another
thing
that
we're
doing
that's
going
to
be
starting
very
very
soon
is
what
we're
calling
speedruns
and
I
think
this
is
being
rolled
out
even
more
generally
than
just
CI
CD.
But
this
is
where
we're
going
to
find
some
basic
use
cases
that
are
either
things
that
you're
doing
all
the
time
in
the
product,
or
are
things
that
you
that
are
really
important
to
new
users
who
are
onboarding
and
just
record
how
long
it
takes
to
do
it.
How
difficult
it
is
where
you
slow
down,
where
you
wait.
A
Things
like
setting
up
a
pages
site
is
a
great
example
of
this,
where
there's
different
weights
and
have
things
that
happen,
and
sometimes
things
are
unclear
and
we're
just
gonna
record
those.
You
measure
the
time
and
measure
how
that's
changing
over
time
and
drive
down
some
key
user
flows
in
the
product.
In
this
way,
if
you
have
suggestions
here
for
or
areas
that
we
should
be
measuring
in
this
way,
do
let
us
know.
A
The
last
of
my
favorite
slide
is
where
we
ask
for
your
contribution
on
different
things,
so
we're
starting
the
product
development
flow
process
on
CI
CD
teams.
As
well,
so
we
may
look
for
your
help
and
connecting
with
customers
if
you're
in
the
support
or
sales
organizations,
and
things
like
that,
we
do
really
need
your
support
in
order
to
get
in
touch
with
good
customers
for
these
issues.
A
So
please
do
help
us.
If
we
reach
out
the
accessibility.
Mbc
is
moving
into
your
problem,
validation
as
well,
so
we're
looking
for
feedback
on
that.
The
container
registry
retention
and
expiration
policy
issue
also
needs
feedback,
we're
trying
to
determine
what
the
best
process
is
and
policies
for
automatically
removing
unused
images.
A
Obviously
you
can
imagine
that's
a
pretty
complicated
topic
and
different
points
of
view
really
really
help
us
we're
also
trying
to
figure
out
how
we
can
integrate
the
current
stage
based
view
and
the
dag
view
of
pipelines
in
an
interesting
way,
and
it
doesn't
result
in
kind
of
like
two
different
views
in
the
product,
and
so
this
is
one
of
the
ones
that
we're
doing
the
product
development
flow
on
but
view
points
here
are
very.
Very
welcome.
A
I
also
have
made
this
Jenkins
migration
doc
a
couple
months
back,
I
think,
and
it
would
be
really
really
great
to
get
more
feedback
from
folks
who
are
working
in
the
field
and
implementing
this
to
refine.
That
document
turn
into
something
really
really
valuable.
I
think
I
saw
that
the
manage
team
or
or
whichever
team
was
working
on
the
importer
is.
He
is
gonna,
be
potentially
using
that
to
base
the
things
that
they're
building
off
of
so
more
insights
into
that
document.
A
Any
feedback,
real
world
experience
is
just
going
to
make
everything
better
for
helping
our
users
who
are
trying
to
get
over
from
Jenkins
onto
our
product
and
then
access
controls
for
pages
enabled
on
get
lab
comm,
which
was
a
great
community
contribution
and
also
love
to
hear
feedback
on
there.
So
with
that,
I
will
stop,
sharing
and
turn
it
over
to
you,
questions.
Let's
see
if
any
have
been
shared,
all
right
go
ahead.
John.
A
So
we
have
an
issue
where
some
of
our
customers-
well,
basically,
everybody
uses,
get
lab.
Comm
is
paying
a
little
bit
of
attacks
where
we're
going
up
to
github
or
sorry
docker,
docker,
hub
and
fetching
containers
every
single
time.
We
were
waiting
on
Puma
to
be
rolled
out,
but
we
have
a
issue
where
we're
trying
to
follow
that
out
via
stage
canary
and
eventually
in
production,
in
order
to
to
test
this
feature
and
bring
bringing
the
capability
out
at
gate,
lab,
comm
scale.
C
I
would
say
we're
dogfooding
in
it
still.
We
still
will
have
to
follow
up
and
do
things
like
add
authentication
for
private
projects.
There
may
be
some
performance
concerns,
but
we're
hoping
that
by
solving
the
dogfooding
problems,
first
for
our
infrastructure
team
that
we'll
be
able
to
drive
more
requirements
from
that
issue
and
prioritize
them
appropriately.
B
A
Yeah,
so
that
one
is
more
in
the
works
as
I
understand,
we
don't
have
a
super
super
finalized
strategy
on
how
that's
being
rolled
out.
Although
and
correct
me
if
I'm
wrong.
This
is
an
important
one,
though
it's
important
for
compliance,
workflows
and
one
interesting
thing
that
we're
doing
is
looking
at
starting
like
a
compliance
at
a
more
cross.
Github,
our
circuit
lab,
get
lab
level,
and
so,
potentially
this
is
going
to
roll
into
that
to
a
certain
extent.
A
D
You
very
much
yeah
Hey
yeah,
we're
potentially
forking
the
docker
registry
I.
Think
I
asked
already
whether
we
can
switch
to
harbor
and
then
already
answered
that
hardware
is
also
the
docker
registry
under
the
hood,
so
that
doesn't
make
sense
and
I
think.
The
reason
for
switching
is
that
there's
kind
of
PRS
on
the
darker
fort
that
aren't
getting
merged
for
Kenyan
open
source
protect
is,
it
is
a
big
statement
and
we
want
to.
D
We
want
to
still
benefit
from
like
the
power
of
open
source
or
other
people
contributing
so
are
we
in
touch
with
docker
like?
Are
we
saying,
hey,
look,
there's
all
these
things
languishing
that
are
not
getting
merged.
We
need
those
face
to
get
merged,
we're
considering
a
fork,
but
we
would
prefer
if
you,
if,
if
you
take
like
get
back
on
maintaining
this
project
or,
however,
we
totally
understand,
if
you
don't
have
the
intention
or
the
research
is
to
do
so,
and
we
are
we
like
opening
that
conversation
in
a
second.
D
If
we
do
have
a
fork,
I
think
we
should
I
can
fight
everyone
else
to
join
us
in
that
fork,
so
that
we're
not
maintaining
for
by
ourselves,
but
that
we
become
kind
of
the
new
active
project.
Maintaining
this,
it
seems
that
other
organizations
like
harbor
are
also
dependent
on
this
I
assume
they
welcome
it.
A
C
So
we
we
have
been
in
touch
with
doctor,
but
only
in
the
issues.
I
think
some
sort
of
reach
out
through
our
alliances
team
and
having
that
conversation,
what
could
be
really
helpful?
I
could
reach
out
to
Brandon
and
see
if
we
can
coordinate
that
we
have
been
in
touch
with
Harbor
a
bit
and
I
think
they
would
be
interested
in
supporting
the
fourth,
because
they've
encountered
these
very
same
problems.
C
They've
been
trying
to
push
changes
to
improve
how
garbage
collection
is
done,
so
I
think
we,
there
is
definitely
an
opportunity
to
work
together
there
and
we
recently
have
just
started
working
through
what
it
means
to
maintain
our
own
Fork.
So
Stan
just
push
the
first
changes
to
improve
error
handling
beyond
that.
Now
we're
starting
to
say
what
does
it
mean
to
maintain
our
own
Fork?
How
do
we
handle
upstream
patches,
like
how
does
all
of
that
work?
So
I
think
we
are
ready
to
go
into
a
conversation
with
both
docker
and
harbor.
C
D
C
Of
our
customers
have
expressed
interest
in
helping
specifically
with
some
of
the
problems
around
garbage
collection
and
I'm
tagging.
They
said
we
would
love
to
partner
on
a
fork,
so
I
think
there's
an
opportunity
there
that
I've
been
kind
of
pushing
back
until
we
made
some
of
these
preliminary
decisions.
That
I
think
we
could
follow
up
on
with
with
those
with
them.
You.
D
So
let's
make
sure
that
we've
done
our
homework.
Also,
let's
make
sure
that
if
we
do
fork
it,
we
do
it
in
a
kind
of
a
low-profile
way.
We
should
hide
anything,
but
darker
has
been
amazing
for
like
contributing
all
of
this
and
then
walking
around
and
like
putting
poor
process.
We're
forking.
This
and
stuff
like
that
is,
is
not
being
a
good
citizen.
D
A
E
F
A
And
so
it's
sort
of
the
two
that
I
touched
on
in
relation
to
the
tomb
where
we
want
to
invest
next
year
and
potentially
pull
those
in
early
is
mobile.
Use
cases
is
a
really
really
important.
One
there's
a
whole
segment
of
customers
that
we're
not
serving
very
well
Windows
and
Mac
runners,
of
course,
but
that's
that's
already
in
the
works,
and
then
onboarding
is
really
the
one
that
I've
become
more
and
more
passionate
about
improving,
just
getting
up
to
speed
the
day
to
day
basic
work,
clothes
and
get
in
get
lab
is.
A
Is
super
super
important
and
there's
a
lot
that
we
can
do
there
and
there's
a
great
issue
that
came
in
yesterday
about
building
the
the
CI
lint
into
the
web,
IDE
more
syntax,
highlighting
all
of
these
things
will
help
people
get
started.
Autocomplete
is
another
great
example.
Is
that
these
that
are
they're
great
ways
to
help
people
understand
the
power
of
get
low?
We've
built
a
lot
of
great
features
that
are
keywords
in
the
CIA
ml,
but
you
sort
of
have
to
read
the
documentation
to
find
them.
A
It's
not
necessarily
automatic
in
a
lot
of
ways
to
to
know
how
to
get
them
set
up.
We
have
auto
dev
ops,
which
can
do
things
to
a
certain
extent,
but
we
want
to
continue
to
expand
that
and
do
more
auto
detection
of
things
more
discoverability
of
features,
and
that's
why
we
want
to
build
a
whole
team.
That's
focused
on
focused
on
doing
that.
I
think
that's
going
to
make
a
huge,
huge
difference
and
really
make
us.
You
know
best
in
the
market
in
a
lot
of
ways
in
terms
of
getting
things
done,.
F
Cool
thanks,
I
had
the
next
one,
which
was
I
know
we
recently
added
direct,
take
directed
acyclic
graphs,
which
enable
you
to
do
different
dependencies
between
jobs
across
stages
and
wondering
if
we
have,
we
seen
any
results
from
that
of
customers
using
it
and
seeing
improved
pipeline
performance.
Yeah.
A
Definitely
I
don't
have
measurements
on
like
percentage
improvement,
but
I've
seen
a
ton
of
activity
in
two
different
ways.
One
is
on
the
issues
or
follow-up
improvements,
one
of
the
great
signs
that
you've
released
a
a
good
feature
that
people
are
really
engaging
with.
As
you
start
getting
a
lot
of
bugs
and
you
start
getting
a
lot
of
feature
requests
around
it
and
that's
been
the
case
with
a
dag
they're,
getting
a
lot
of
uploads
quickly.
A
We've
got
strong
signals
that
we
want
to
continue
investing
in
this
and
making
the
feature
better
and
better
and
then
the
other
area
that
another
place
that
I'm
seeing
it
is
on.
Twitter
people
are
sharing
that
you
know
like
wow,
you
know,
I
just
used
the
needs,
keyword
and
parallel
keyword
and
was
able
to
improve
my
pipelines
and
I
always
find
in
and
retweet
those
and
really
enjoy
seeing
this.
A
So
definitely
seeing
that
this
was
a
really
good
investment
and
once
we're
able
to
do
that,
research
to
figure
out
how
to
unify
the
dag
and
the
the
current
stage
based
model
I
think
it's
gonna
be
we're.
Gonna
have
a
really
unique
solution
on
the
market,
because
nobody
else
has
this
hybrid
view
that
we
that
we
have
where
you
can
use
the
stage
based
model
independently.
You
can
use
the
dag
independently
or
you
can
use
both
at
once
and
kind
of
stage
things
out
within
different
stages.