►
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).
B
So,
let's,
okay,
so
probably
you
saw
the
developer
survey
that
we
just
did
and
the
thing
that
struck
me
the
most
was
that
no
matter
who
you
talk
to
in
the
developer
survey
developers,
security,
people
or
ops
pros
you
got
the
same
answer.
50%
of
all
of
them
said
test
is
the
biggest
problem.
When
it
comes
to
slowing
down
software
development,
I
mean
they
just
that's
it,
they
all
dumped
on
tests.
So
let's
start
there.
Why
is
this?
Does
this
seem
to
be
a
continual,
a
continued
DevOps
problem
that
everyone
struggles
to
to
solve
so.
A
I,
don't
categorize
that
into
buckets
first
just
to
set
them
the
view
of
the
high
level
I.
Think
the
first
problem
is
that
they're,
not
enough
tests,
that's
one
main
thing.
The
second
is
there
are
tests,
but
they're
flaky,
ok
and
let's
start
with
the
first
bucket
first,
where
there
are
not
enough
tests,
and
this
is
also
a
challenge.
We
are
addressing
you're
a
kid
laugh
I
feel
after
say,
I
think
we're
better
than
other
companies
where
developers
write
unit
tests
and
integration
tests
every
time
a
change
goes
in.
That
is
great
with
that.
A
Testing
is
at
a
really
low
level
and
it
doesn't
really
match
to
a
business
use
case,
and
we
rely
on
this
on
business
requirements
to
come
in
product
management,
spread
out
the
information
to
the
group.
That's
working
on
so
UX
back
in
front
end
tests
correct
and
these
three
four
sets
of
stakeholders
need
to
understand
what
is
the
end
goal
and
then
the
requirements
can
be
defined
and
then
test
requirements
which
isn't
really
being
used
as
often
like.
There
are
part
deployments
but
hey.
A
We
need
test
requirements
to
write
tests,
so
that
facet
is
missing
and
we're
improving
it
here,
I
get
lab
where
I
think
it's
got
is
doing
great
job
BPO
product
I.
Think
the
day
way
of
doing
things
is:
if
you
have
a
product
requirement
you
need
to,
we
need
to
define
the
test,
deliverables
or
the
test
requirements.
We
have
a
section
for
it
right
now,
but
it's
the
template
it's
blank,
so
we
it's
a
free
form.
A
People
can
fill
in,
but
it
should
you
highlighted
going
forward
as
a
thing
so
going
back
the
stakeholders
who
are
delivering
this
needs
to
understand
the
end
goal
right.
A
unit
tests
is
testing
code
at
a
smaller
scale,
it's
great,
but
it
doesn't
really
underscores
or
verifies
that
the
function
e,
what
functionality
works
and
to
end
it
alone.
So
we
need
more
coverage
and
more
understanding
of
what
needs
to
be
tested.
I
see
that
it's
like
the
first
enablement
part
of
lobbyists
are
not
enough
tests.
A
Another
factor-
and
that
is
the
test
infrastructure
unit
test-
is
great.
It's
easy
to
set
up,
but
we
need
a
team
that
has
that
has
the
specialties
and
knowledge
that
builds
up
this
end-to-end
test
infrastructure.
It's
a
lot
higher
than
writing
unit
tests.
You,
justice,
I,
wrote
this
function,
I'm
writing
a
test
that
pairs
with
it
and
it
just
test
a
function.
That's
great
we're
doing
it
at
the
low
level
making
sure
all
the
the
gears
are
round
and
in
good
shape
before
you
fit
it
in
correctly.
A
A
A
lot
of
organizations
will
have
end-to-end
tests,
including
us
as
well
we're
trying
to
improve
that
and
do
it
has
comes
later,
like
the
ideal
goal
is
like
when
things
starts
to
shift
or
starts
to
get
built
right.
The
tests
developments
need
to
start
at
the
same
time,
and
then
you
assemble
it
together
run
it
it's
passing.
We
go
for
launch
and
it's
shipped
we're
not
there
yet
and
I
can
assure
you
a
lot
of
companies
out.
There
are
not
there
yet
as
well.
A
This
is
like
the
ideal
goal
where,
where
should
be
setting
our
true
compass
on
now,
that's
the
first
part.
The
second
part
is
the
flaky
tests.
Psyche
tests
and
I
would
even
tack
on
the
ramp
up
time
to
write
a
test
there.
As
well
so
essentially
making
the
framework
and
mature
and
stable
I
think
there's
very.
How
do
I
say
this?
Not
their
love
test,
automation,
engineers
and
tests
developers
out
there,
but
not
all
them
know
how
to
write
and
design
a
good
test.
What
I
mean
is
I?
Compare
this
with
my
team
is
falling.
A
The
test
needs
to
be
like
a
cell-free
trying
dumb,
you
know,
mm-hmm,
like
in
every
review,
are
being
useful
floor
work.
We
abstract
it
to
like
a
click
method,
draw
method,
I,
get
element
method
and
all
these
needs
to
have
some
fault,
tolerance
and
retry
ability
in
them
like
when
you
look
at
an
a
manual
tester
per
se.
Nobody
clicks
on
a
button
and
then
just
wait.
Ten
minutes
and
watch
it
to
pass
right.
If
he
clicks
on
the
button
doesn't
go,
he
just
clicks
it
again.
A
You
know
like
these
flakiness,
like
hey,
if
you,
if
you
click
on
it
and
there's
some
latency,
your
core
framework
with
this
abstraction
levels
needs
to
be
able
to
retry
it
and
like
using
as
human
centric
approach
like
I.
Don't
see
anybody
being
putting
this
emphasis
on
the
design
of
the
framework
we
are
I
think
we
call
it.
Stateless
page
objects
where
the
retry
ability
needs
to
be
picked
in
the
retrying
domino,
because
you
have
a
user
workflow.
A
If
that
Domino
doesn't
go,
he
needs
to
make
sure
the
next
guy
goes
so
the
workflow
is
completed.
A
lot
of
tests
would
just
be
like
at
one
point
feeling
we're
okay
fail
and
then
it
doesn't
even
execute.
That's
that's
one
example
tests
that
are
too
complex
like
making
sure
the
tests
are
laid
properly.
A
Video
is
this
magically
appeared
at
the
mr
page
because
you
don't
have
to
go
through,
like
you
just
bypass
the
whole
mechanism.
A
lot
of
bad
practices
is
to
do
everything
through
UI,
where
you
testing
the
it
more.
But
then
you
spin
up
the
brow,
so
you
type
the
login
press
lock
in
you
go
to
this
page.
You
go
to
two
more
pages,
and
then
you
finally
reach
the
mr
page.
Where
that
test
is
just
needs
to
test
the
mr
page,
you
don't
have
to
test
the
rest
of
it
right
so
like
focusing
the
test.
A
B
Awesome
analogies
and
by
the
way
the
Apollo
11
one
is
awesome.
That
was
great
because
I
had
never
thought
of
it
in
that
way
before,
so
that
that's
awesome.
So,
but
let's
talk
about
this
whole
developer
involvement,
it
always
seems
like
there's
a
little
bit
of
friction
about
who
owns
what
I
mean
how'd
it
happen.
How
can
I,
how
can
a
team
kind
of
work
through
that
and
get
to
this
sort
of
ideal
of
everybody's
really
responsible,
because
it
seems
like
so
much
of
software
development
is
handing
off
right.
So
can
we
talk
about
that?
B
A
Sure
I
think
it's
part
of
the
DevOps
transformation
is
everybody's.
Saying,
like
this
team
owns
everything
I
think
there
are
many
ways
to
achieve
this.
For
anything,
we
set
the
direction
that
every
but
quality
is
everybody's
responsibility.
I
actually
saw
a
recording
thing.
It
was
Tesla
I
think
when
they
walk
through
the
factory
and
that's
like
a
banner
in
quality's
everybody's
responsibility
up
there
I
think
we
do
a
good
job
of
that,
but
setting
discussions
and
enabling
the
team
to
achieve
that
goal.
A
The
suit
needs
to
be
improved
and
I
would
imagine
that
the
same
challenge
the
other
teams
were
doing
as
well.
There
are
many
ways
likely
they're
been
at
first,
you
have
developers
be
on
call
where
you
were
responsible
for
your
code
when
you
gosu
production
that
is
kind
of
a
mechanism
to
go
there.
Behaviorally
we
want
to
encourage.
Is
that
hey
when
you
write
this
code,
do
in
such
a
way
that
you'd
have
to
stay
on
Saturday
to
make
sure
that
it's
my
codes,
gonna
pass
or
not
right,
pull
in
your
stakeholders
like
ask
quality?
A
It's
this
enough
testing,
the
Oscar
p.m.
maybe
cover
all
the
cases
and
a
with
the
bringing
up
of
DevOps
transformation.
I,
don't
think
all
the
developers
have
that
mindset
baked
in
yet
I
think
it's
getting
there
but
like
having
people
think
about
these
repercussions
and
how
to
mitigate
them
earlier
on
and
and
not
just
focusing
on
shipping.
A
The
features
I
think
it's
key
I
think
I
think
it's
a
concerted
effort
per
se
thing
and
there's
all
the
extreme
where
like,
for
example,
I,
don't
want
to
name
companies
but
they're
companies
that
will
call
engineers
on
the
Saturday
night
at
1:00
a.m.
and
then
and
then
they
leave
it
there
where,
like
that's
like
the
means
to
achieve
that,
but
they
don't
really
take
that
feedback
in
to
improve
the
process.
A
If
that
doesn't
happen
in
the
future,
it's
like
okay,
the
engine
is
gonna
fix
it
on
Saturday,
we're
good
we're
happy
it
fixes
on
Sunday.
Our
customers
are
happy,
but
they
can
point
out
the
companies
that
treat
engineers
like
resources
and,
like
everybody,
does
everything
and
then,
if
people
get
born
out
the
door
is
there
we
can
hire
more
because
we
have
so
much
money
in
the
bank,
we're
a
public
company
and
that's
not
where
I
want
a
good
laugh
ago.
A
I
think
I
I
want
to
see
people
be
happy
at
their
jobs
and
the
pathway
that
they
need
to
be
set
should
be
that
we
we
show
people
where
it's
going
wrong.
Let's
make
the
process
better,
so
we
work
smarter,
achieve
more
without
having
to
like
put
down
our
weekends
or
like
get
ping
doing
your
families
in
dinner.
Nobody
wants
that.
I
I
can
see
why.
A
Folks
that
didn't,
transform
to
DevOps
properly
will
fall
back
to
the
old
as
a
way
of
having
a
wall
where
you
have
developer
in
QA,
and
here
we
just
test
everything.
This
is
because
then,
the
eliminates
the
on
call
aspect,
because
there's
a
whole
department
just
doing
testing
for
you
and
that's
like
going
back
to
like
you
and
you're,
just
like
a
reversing
or
removing
progress
at
that
time,
and
it
was
but
then
then
you
need
140
engineers
testing
right.
It.
A
It
works
and
the
disciple
time
is
flawed
because
the
wall
is
there,
he's
gonna
get
tested,
but
it's
gonna
take
a
quarter
to
like
release
something
and
that's
not
where
we're
headed
when
it
release
fast.
So
in
order
to
get
to
that,
then
we
need
to
improve
the
process
and
then
make
people's
lives
easier.
It
works
water
speaking.
B
Of
working,
smarter
and
making
people's
lives
easier.
Let's
talk
about
test
automation
because
it
seems
like
that
again
in
talking
about
DevOps
or
devstack
ops
or
scaling
DevOps
test
automation
is
so
crucial,
but
yet
everybody
struggles
to
kind
of
get
there.
So
is
it?
Are
we
waiting
on
new
technologies
like
AI
or
ml,
or
are
we
you
know?
What
do
you
think
we're?
Where
is
what
is
test?
Automation
really
gonna,
be
like
oh
yeah,
that's
just
it's
just
gonna,
be
there
like
when's,
it
gonna
be
I,
guess
maybe
I'm
asking
less
painful.
A
A
This
is
something
that
I've
done:
Tech
Talks
long
before
the
selenium
conference
in
sorry
that
Stila
to
meet
up
inside
this
go
you
need
to
design
a
product
so
that
the
test
automation
framework
can
integrate
to
it
and
like,
for
example,
if
you're
testing
a
page,
the
gold
standard
is
selenium,
which
is
if
the
text
a
locator
correct.
We
have
a
good
practice
circuit
lab,
but
we
have
QA
locators
like
when
people
see
this
guy
is
being
used
for
test
automation.
Nobody
change
it
until
until
the
test
has
changed
as
well.
A
Right,
like
the
testability
part,
when
other
teams,
where
folks
on
care
about
testability
the
framework
means
to
attach
this
locators
to
whatever
the
locate
locator
it's
available,
and
sometimes
it's
random
it
just
it's
auto-generated.
Every
time
two
page
renders
that's
like
the
the
worst
case,
where
I've
been
in
places
where
the
index
will
refresh
or
there's
like
a
wild
card
like
five
characters,
but
it
would
change
every
time
you
refresh
the
page.
So
you
need
to
make
work
around
this
logic
in
your
framework
to
like.
A
Okay,
I
have
to
detect
this
and
then
what's
gonna,
be
the
next
string
that
gets
refreshed
next
time.
Is
this
this
locator
rendering?
So
that's
that's
a
lot
of
hardship
to
make
them
mm-hmm
the
framework
attaches
so
well.
That's
the
UI
aspect.
The
backend
aspect
is
providing
API
endpoints
to
set
up
workflows
or
set
up
test
dependencies.
Like
I
mentioned
the
login
page
before
we
comparing
to
get
lab,
it
could
be
like
I
need
to
merge.
A
Building
these
api's
require
collaboration
with
development,
cuz
you're,
adding
code
to
the
back-end
code
to
have
this
mechanism
or
a
door
or
an
interface
whenever
term,
for
this
test
framework
to
integrate,
I
think
you
go
back
to
the
Apollo
and
it's
like
these
are
like
the
holes
and
it
Rockets
like
I
need
to
integrate
with
this
image
with
these
and
what
you
find
that
probes
and
sensors
need
to
be
there.
A
So
if
it
goes
on
there,
then
you're
essentially
hitting
the
wall
and
it's
not
working
smart
right,
you're
taking
the
log
around
for
the
test,
automation
engineers
to
cool
it
around
this,
and
if
you
were,
if
it's
there,
yeah,
that's
and
and
it's
not
by
email.
We
do
have
some
discussions
here
and
good
lab
work.
We
want,
but
I
think
ml
will
come
and
help,
but
the
input
and
output
needs
to
be
clearly
defined.
So
you
can
take
advantage
of
the
machine.
A
Learning
intensive
well
regression
or
whatever
the
the
stainless
page
object
that
I
mentioned
to
you
and
that
we
trying
Domino
is
an
or
piece
of
that
like
when
you
can
extract
your
framework
in
such
a
way
that
it's
just
lower
level
instructions,
I
click.
This
click
this
with
this
and
those
are
stable.
You
can
write
a
bot
that
just
lives
in
the
product,
meaning
like
you
can
have
a
bot
that
looks
at
a
page.
He
looks
at
all
the
key
male
okay,
yes,
okay,
what
can
I
click
today?
A
First,
like
let's
let
these
things
run
through
Crawford,
the
application
where
the
hot
spots
and
then
I
structure,
it's
such
a
way
that
he
is
like
the
the
really
really
menacing
hue
engineer.
Where
I
failed.
Is
this
today
I'm
going
to
go
there
again
and
fail
at
the
end
they're
getting
failed
again
and
then
the
fourth
floors
have
passed
like
I'll?
Go
there
next
next
day
or
next
month,
but
like
a
this,
is
failing.
So
let's
go
there?
Okay,
so
they
I
think
that's
the
structure!
B
Cuz
I
think
that
has
a
lot
of
potential
I
think
it
sounds.
It
sounds
cool,
so
hey!
Here's,
a
random,
slightly
random
question.
So
when
we
did
the
developer
survey,
one
thing
that
came
through
loud
and
clear
was
that
73%
of
companies
do
absolutely
no
accessibility.
Testing
like
none.
Is
that
surprise
you
it.
A
We
actually
have
issues
attract
that
like
we
don't
we
want
to
do
automated
accessibility,
testing
color
for
the
color
blind
per
se
like
these
needs
to
be
brought
up
more
fun
and
center,
but
given
the
priorities
right
now
other
startups,
they
will
focus
on
getting
a
functionality
to
work
first,
and
these
will
complete
it,
and
it's
the
same
with
localization
testing
I
will
do
in
fingered.
Localization
has
has
a
higher
priority
sation
than
excess
ability,
mm-hmm
I.
Think
that's.
That's
probably
why?
Okay.
B
That
makes
sense
cuz
like
that
number
just
was
stunning
to
me
like
seriously.
Cuz,
it
seems,
like
you
know,
should
be,
as
you
say,
with
sort
of
the
whole,
it
should
be
more
inclusive
okay.
So
if
you,
where
do
you
see
what's
next
for
testing
it
get
lab
or
or
what
what's
next
or
what
can?
What
can
people
start?
Thinking
about
I
mean
what
are
you?
Are
you
hearing
stuff
from
customers
about
yeah?
We
really
want
this
or
really.
A
A
Have
a
big
check
box
under
enterprise
features
every
time
we
deploy.
We
think
we
do
a
good
job
at
the
at
the
unit
tests
layer,
but
it
needs
to
be
better
because
there's
some
educators
and
bugs-
and
they
especially
you,
know,
enterprise
flow.
We
would
want
to
have
this.
In
first
and
my
team,
my
department
is
working
on
it,
biz
quarters
to
have
a
full
suite
of
enterprise
features
and
doing
to
us.
It's
not
like.
A
We
don't
want
to
add
more
to
the
test
and
run
time,
but
we
need
it
because
this
is
like
mapping
to
the
business
use
case.
You
know
all
the
enterprise
feature.
What
whole
that
needs
to
be
certified.
Have
it
be
a
thing
for
every
deploy
like
every
deploy?
We
verify
that
this
works.
Any
parts
customer
should
be
happy,
I
think
that's
like
the
next.
The
next
cover
actually
proven
that
we
will
do
the
next.
One
is
productivity
where
you
need
to
do
all
that,
while
shortening
the
test
run
time
for
our
developers.
A
It's
it's
kind
of
like
a
self-balancing
gaming.
You
want
more
test
coverage,
but
we
need
to
keep
the
runtime
low
because
we
can't
have
developers
wait
two
hours
right,
so
adding
more
runners,
optimizing
them
removing
some
tests
and
make
sure
that
it's
shortened
down
I
would
even
say
I
would
like
to
see
30
minutes.
It
may
be
a
really
aggressive
goal
for
this
year,
but
right
now,
it's
like
an
hour
or
so,
but
it
would
love
to
have
it
down
to
30
it's
where
we
certify
that
this
amount
going
in
checks.
A
B
A
Productivity
is
a
term
that
has
recently
come
up
in
the
industry.
Google
no
longer
calls
their
engineers
QA
engineers,
they
call
them
software
engineer
in
test
similar
structure
or
toss
software
engineers
in
tools
and
tests,
infrastructure,
I,
think
Facebook,
I,
don't
know
what
what
the
term
is,
but
essentially
it's
quality
Department,
but
this
all
is
to
increase
productivity
by
having
this
test
framework
being
available
for
people
to
use
it's
easy
to
use
and
we
provide
metrics
like
test
coverage.
A
A
I,
don't
think
we're
diet
lab
because
for
us
to
transform
that
way,
it's
in
a
period
of
years,
because
it's
a
cultural
change,
there's
a
skill
set
update
right
and
it
means
that
all
developers
can
I
know
where
the
test
gaps
are
and
the
requirements
are
clear
to
them
that
you
need
to
be
enabled
and
then,
with
the
tools
and
productivity
tools
they
they
can
just
run
with
it
on
their
own
with
us.
Taking
a
step
back
and
maintaining
the
infrastructure
is
solid.
A
Would
have
one
team
so
the
way
we
do
it?
A
good
lab
right
now
is
that
we
still
maintain
a
test,
automation,
quality
counterparts,
but
we
have
a
separate
team
that
takes
on
this
work,
though
it's
a
smaller
team
and
that's
how
we
will
over
time,
it's
not
like
we're
not
looking
to
be
are,
but
there's
a
line
of
separation
of
specialists
right
now
in
the
future.
A
I
do
not
know
it
could
be
that
these
two
roles
are
merging
until
we
do
one
that
also
comes
up
with
title
changes
and
whatnot,
but
there
are
other
companies
that
have
transformed
this
already
and
I.
Think
a
good
example
is,
is
Google
Google
cloud.
Actually
they
do.
This
type
of
home
engineering
productivity
eat
enable
mine
work,
oh
cool.
B
All
right,
well
so
listen
you'll,
be
hearing
from
me
next
week.
I'll
send
you
I'll
write
up
this
blog
post.
You
can
change
it
edit,
it
whatever
and
then
we'll
take
it
from
there,
but
thank
you
so
much
for
the
time
and
a
little
taste
of
California,
because
hey
I'm
home
sick
and
have
something
like
fabulous
and
ethnic
for
me
because
living
north
of
Boston.
It's
a
lot
of
boiled
dinner
here
and
seafood
and
not
like
a
lot
of
other
fabulous
stuff
so
and
I
happen
to
know.