►
From YouTube: Verify 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
A
B
A
B
Yeah
and
I
seen
this
a
few
times,
but
we've
we've
ended
up
with
pretty
ugly
URLs
and
things
like
that
because
we
were
like.
Oh,
we
didn't
reserve
that
thing
and
we
did
the
shortest
thing
that
it's
gonna
for
that
for
the
next
20
years.
That
means
the
things
I
are
not
gonna
be
as
the
way
they
should
be.
A
I
know
we're
gonna
be
careful
with
the
things
that
we
deprecated
in
a
minor
release,
but
I
do
think.
We've
got
that
functionality
built
into
how
we're
looking
at
releases
with
the
major
releases,
but
also
you
know
we
do
some
kind
of
time.
Deprecated
things
mid,
mid,
major
I
guess,
isn't
a
minor
releases
so.
B
A
There
really
gets
to
the
core
of
the
iteration
value
to
not
just
that
we're
iterating
the
functionality
but
we're
iterating,
saying
like
hey
the
way
we
thought
this
through
with
the
first
time
after
we
learned
about
how
it's
being
used
and
like
how
we
can
build
on
it,
we
need
to
kind
of
tweak
that
a
bit
too
and
and
change
it
a
bit
so
I
completely
agree.
Yeah.
B
And
it's
an
in
case:
it's
like
this.
It's
already
clear
about
the
preferred
ways.
We
just
can't
go
there,
so
we're
gonna
ship,
the
value
to
customers,
get
it
get,
it
done,
provide
the
value,
and
then
we
have
a
kind
of
a
cleanup
task.
That
is
it's
almost
a
it's,
not
technical
depth
as
much
a
huge
interface
that
or
URL
URL
depth.
Where
you
were
you,
you
want
to
change
that
someone's
gonna,
correct
me
and
say
you're
right
amped,
something
like
that.
B
B
A
C
Yeah
I
can
talk
about
that.
So
there's
a
few
things.
We
we've
pushed
this
back
a
release.
Then
we
wanted
to
do,
but
we're
gonna
be
doing
a
proof-of-concept
first.
So
today,
it's
possible
to
integrate
ball
with
gitlab
the
same
way
that
ball
integrates
with
Jenkins,
which
is
through
an
off
method
that
they
call
a
Crowell.
It's
basically
a
single
factor,
server-to-server
off,
which
you
could
do
today
with
a
get
loud
runner,
but
we
don't
want
that
to
be
the
only
way
to
do
it.
C
So
we're
looking
at
integrating
TLS
off
the
authentication
directly
into
the
runner
so
that
the
runner
can
receive
like
a
temporary
key
from
vault
do
its
deploy
task
or
whatever
other
task
it
has
and
then
and
then
that
key
expires
after
a
certain
period
of
time.
You
know
ten
minutes,
let's
say
so
that
that's
what
we're
working
on
currently
and
then
I
think
that
from
that
we'll
learn
more
about
how
we
might
be
able
to
kind
of
then
there's
also
a
migrating
to
vault
epoch.
That
comes
after
that.
C
That
was
mostly
a
set
idea,
so
I'm
explaining
as
I
did
in,
but
the
idea
that
you
know
secrets
management
in
general
and
get
lab
forget
just
verify,
but
but
throughout
the
application.
It's
something
that
we
should
should
consider
is
vault
a
better
solution
than
trying
to
iterate
on
our
own
secrets
and
keys
that
we
use
within
the
application
and
so
that
that
is
kind
of
where
we're
headed
but
we're
at
the
very
beginning
of
it.
So
hopefully
I
think
it.
So.
The
question
said:
yeah.
B
B
C
It's
not.
We
haven't
made
that
decision,
yet
no
I
think
we
wanted
to
go
through
the
proof
of
concept
before
we
decided
to
do
that.
I
think
it's
very
likely
outcome
that
we
would
consider
that
right
because,
for
instance,
we
just
shipped,
you
know
that's
what
we
called
simple
masking
of
variables
right,
which
is
a
very
simplistic
way
of
not
really
protecting
the
variables,
but
still
something
customers
wanted
to
mask
them
and
the
build
logs.
C
Whereas
again,
the
more
mature
way
of
handling
that
would
be
to
have
variables
that
are
immutable
and-
and
you
know
sorry
secrets
that
are
amenable
and
go
away.
So
I
think
that
we
will
probably
eventually
say
that
that
is
the
recommended
method.
But
I
I
want
to
have
a
better
understanding
of
how
tight
of
an
integration
we're
going
to
be
able
to
achieve
before
we
decide
hey,
let's
scrap
our
variables
completely,
for
it.
B
Yeah,
it
makes
sense
to
iterate
what
I
don't
understand
is
like
I
see
issues
like
this.
It's
to
insert
install
hash
occult
onto
the
cluster.
That's
attached
to
get
lamp
instead
of
like
the
cluster
or
omniverse
situate
that
Gil
app
is
running
in
so
this
would
mean,
for
example,
for
get
Lancome.
We
have
thousands
and
thousands
of
people
attaching
clusters,
we
would
be
running
thousands
and
thousands
copies
of
false
I
thought
that
the
idea
was
that
vote
would
become
part
of
get
lab
and
it
would
be
managed
next
to
to
get
lab
installation.
B
So
there
would
be
a
forget.
Lokhande
would
be
one
revolt
installation
with
all
the
secrets,
because
I
get
lab
environmental
variables,
environmental
secrets
are
handled
in
get
lab
not,
and
it's
not
a
it's,
not
a
deploy
target.
It's
it's
the
essence
of
the
application,
like
we
have
a
single
database
for
get
Lancome,
etc.
Why
should
I
see
that
yeah.
C
B
That
makes
sense
in
a
like
the
verbal
way.
You
explain
it,
however,
from
reading
this
issue,
I
couldn't
I
couldn't
get
that
context
and
I
would
I
would
think
that
looking
at
the
issues
or
company
strategy
is
to
treat
vault
like
we
treat
from
meteors
or
something
which
we
expect
to
be
set
up
on
the
deployment
target,
this
isn't
screaming
at
me,
hey
fall,
it's
gonna
be
an
essential
part
of
get
lamp
or
there's
two
alternative
routes
and
we're
gonna
pick
one
of
them
sure.
C
Yeah
I'm
gonna
update
that,
because
I
think
that
there's
focusing
on
those
two
silos
I
mentioned
has
thus
then
made
it
confusing
for
you
to
see
the
overall
thing
right.
Link
use
in
so
I'm
gonna
update
the
original
issue,
which
is
super
popular
right
and
has
a
lot
of
input
with
links
to
the
two
epics
and
discussion,
the
Bell
about
the
migration
integration
strategy.
Oh
I
think
that
as
a
takeaway
thanks
for
that.
C
B
How
do
we
I
saw
that,
where
we're
making
progress
towards
like
a
windows
executor
for
the
docker
containers?
How
is
that
going,
and
then?
Oh,
maybe
also
is
this
related
to
like
running
iOS
built-in
docker
like?
Is
that
also
a
direction
we're
going
in,
but
maybe
start
with
that
the
row
to
Windows
testing
the.
A
A
The
kind
of
the
the
biggest
example
of
something
that
has
come
up
and
has
been
a
bit
of
a
challenge
is
just
look
the
way
that
things
are
different
when
you're
running
Windows,
Windows,
docker
inside
windows,
and
so
in
that
case,
the
two
versions
of
Windows
the
host
and
the
client,
their
terminology.
Negatives.
The
hosts
in
the
client
need
to
be
running
the
same
version
of
Windows,
exactly
which
is
a
very
different
sort
of
ecosystem
than
we're
used
to
if
you're
running
docker
with
things
like
Linux.
A
B
A
B
C
I
am
I'm
aware
of
it.
We
haven't
done
any
discovery
rather
yet,
but
we
are
going
to
do
this
cover
around
it.
I
know
that
you
know
Jason
did
a
lot
of
great
work
in
the
first
quarter
around
fast
lane
and
and
builds
for
iOS
and
Android
around
that.
But
there
is
a
growing
push
kind
of
away
from
that
and
towards
this
new
way
to
build
iOS
on
Linux
and
so
I.
Don't
know
a
lot
about
it.
I
just
know
that
it
exists.
B
B
B
B
C
So
we
would
love
feedback
on
that
and
then
I
think
that
merge
request
will
help
us
communicate
better
to
the
outside
world.
That
pay
like
we're
off
track,
and
this
is
the
new
plan
and
so
that
that
covers
all
of
the
the
kind
of
C
ICD
stages
which
are
verified
package
and
release
from
our
time.
Stitches.
B
C
C
B
C
B
Because
I
was
there's
a
quote
on
the
slide
number
three,
a
lot
of
types:
people
don't
know
what
they
want
until
you
show
it,
which
can
be
true,
but
also
we're,
as
a
company
I
think
we
have
to
now
get
back
a
bit
more
to
listening
to
customers
and
don't
necessarily
make
what
they
asked
for,
but
make
sure
we
solve
their
problem.
Yeah.
C
I
agree,
I,
think
and
I
think
that
that's
a
balance
that
we're
trying
really
hard
to
achieve
right,
especially
in
verify
right.
We
have
a
loveable
category
in
CI,
but
one
that
we
need
to
maintain
that
leadership
with
well,
of
course,
also
shipping
towards
the
breath
vision.
So
thanks,
ed,
yeah,
I
came
back
marks
voice
may
be
lost
from
the
the
great
time
we
had
to
go
to
New
York
with
our
customers.
The
customer
advisory
board
meeting
was
put
together
by
product
marketing,
but
I
kept
saying.
C
C
So
I
I
had
a
lot
of
folks
interested
in
you
know.
I
talked
about
a
couple
of
things
that
the
Jason
and
I
have
been
talking
about,
so
I
dia
of
progressive
delivery.
If
anyone
doesn't
know
what
that
is,
I
recommend
reading
the
the
C
ICD
vision
page,
and
so
this
idea
of
even
moving
past
continuous
delivery
to
to
what
some
in
industry
are
calling
progressive
delivery.
I'm
Jason
wrote
a
lot
of
great
stuff
about
that,
so
some
of
our
our
largest
customers,
you
know,
see,
have
these.
C
These
teams
that
are
pushing
the
boundaries
of
how
fast
can
they
shift
into
production
and
how
well
can
they
do
continuous
delivery,
even
in
the
regulated
environment,
and
so
while
they
may
have
many
teams
that
they're
still
kind
of
bringing
along
and
bringing
in
to
DevOps
even
or
they
have
some
teams
that
are
really
pushing
the
boundaries.
So
this
kind
of
spectrum
of
customers
that
we
see
of
maturity
along
the
DevOps
lifecycle
and
their
DevOps
maturity
that
same
spectrum
exists,
then
within
many
of
our
customers
too.
C
So
it
was
an
interesting
thing
to
discuss
with
folks
and,
of
course,
the
folks
that
are
there
and
invested
in
get
lab
and
and
showing
up
or
are
helping
to
lead
those
teams
that
are
kind
of
on
the
cutting
edge,
but
they
they
want.
You
know
to
make
sure
that
we
provide
tools
and
ways
to
help
them
get
the
rest
of
their
team's
kind
of
on
board.
So
that
was
that
was
pretty
interesting.
C
A
A
Everyone
started
to
drop
off,
so
let's
do
it,
let's
wrap
it
up.
People
a
couple
minutes
before
the
company
call
thanks
everyone
for
the
questions
and
comments
and
read
through
the
slides.
If
you
have
any
other
questions
or
comments
afterwards,
we
are
in
the
G,
underscore
verified
channel
and
feel
free
to
jump
in
there
and
ask
us
anything
every
day.