►
From YouTube: Quality Department Watir Pageobject Library demo and AMA
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
Hey
everyone:
this
is
the
kinla
kim
lab
ama,
slash
demo,
where
dan
will
be
having
a
recap
of
the
the
water
page
object
architecture.
That's
now
in
a
hybrid
state,
with
our
existing
test
framework
in
the
main
main
repo
that
will
show
us
a
brief
overview
and
then
we'll
go
back
into
questions
to
follow
on
the
the
great
demo
that
we
had
in
the
department
called
a
week
and
a
half
ago.
B
I
am
bundling
chem
lab
inside
of
actually
just
a
recap
of
chem
lab
so
right
now
we
use
capybara
for
our
for
our
test,
writing
and
kind
of
the
back
end
or
the
the
background
for
our
automated
testing.
It
was
intended
long
ago,
a
couple
years
ago,
at
least
of
sort
of
a
migration
to
water,
for
several
different
reasons
which
we
can
cover
in
the
ama.
B
The
intention
is
for
chem
lab
to
be
sort
of
a
a
uniform
directive
for
quality
to
to
have
a
a
way
to
say,
for
instance,
if
you
are
in
create-
and
you
want
to
go
to
plan
you
could-
which
we
can
right
now,
but
if
you
wanted
to
go
to
customers.you'd,
have
to
relearn
an
entire
entirely
new
framework.
So
the
effort
here
is
to
sort
of
join
how
we
do
things
at
ulab
in
our
automation,
as
well
as
customers,
dots
automation,
because
we
have
a
lot
of
similarities
but
they're.
B
B
B
B
So
that's
sort
of
the
the
base
difference
between
page
library
and
page
element,
so
say,
for
instance,
with
that
text
box
you
wanted
to
get
the
text
from
that
element
like
the
value
was
filled
into
it.
Instead
of
adding
a
new
method,
you
just
call
the
method
instead
of
instead
of
defining
a
whole
new
method,
and
you
could
see
that
in
scale
with
say
tomorrow,
we
needed
to
add
something
to
click.
The
text
box,
for
whatever
reason,
we'd
have
to
actually
put
an
element
or
sorry
define
a
new
method
to
collect
that
element.
B
So
much
like
we,
we
define
the
elements,
we
define
the
element
type
and,
if
you
recall,
from
our
element
naming
convention,
we
have
underscore
button
underscore
field
underscore
content,
that's
sort
of
all
managed
within
the
the
tag
itself,
so
text
field
and
button.
So
it
kind
of
removes
the
necessity
for
that
superfluous
suffix,
like
continue
button.
B
B
B
So
this
file
is
the
page
trials
new
which,
as
it
appears
right
now
in
the
in
the
merge
request
yep.
I
have
to
do
here,
but
let's
see
so,
we
have
two
pages
within
the
page
objects
right
now
and
they
are
directly
adjacent
to
all
of
the
other
libraries
or
sorry
page
objects
that
we
have
so,
for
instance,.
B
B
This
is
where
we
call
the
the
copy
bar-
I'm
sorry,
the
chem
lab
page
library
we
perform
it
just
like
we
do
with
the
page
objects,
but
we
have
the
setters
right
here.
So
company
name
equals
what
this
does
is
sets
the
text.
So
we
don't
have
to
do
like
a
new
dot
set
company
name.
We
don't
have
to
do
new
dot,
set
number
of
employees,
set
telephone,
number,
etc,
and
so
yeah
and,
like
I
said,
back
to
back
this
sort
of
runs
in
tandem
with
our
chem
lab.
B
B
A
Then,
where,
where
are
the
methods
defined,
so
it
looks
like
we're
powering
the
the
page
object
now
called
page
library
by
the
type
of
elements
with
the
methods
to
interact
with
the
locators,
like
click,
the
the
click
the
drag
in
the
future,
find
elements
how
how
are
they
being
plugged
in
again?
If
you
can
show
that
again,
please.
B
Yeah,
so
I
can
actually
quickly
go
over
without
showing
it,
but
basically
it's
held
within
a
rubygem
called
chem
lab
and
what
you
just
referred
to
mac
is
the
w3
standard
of
sort
of
the
actions
you
can
perform
on
specific
elements.
Those
actions
are
defined
in
chem
lab,
so
you
don't
have
to
worry
about
those
sort
of
actions
as
new
like
say,
the
w3c
standard
iterates
and
adds
a
new
action.
We
just
update
kemlid
with
that,
and
we
don't
have
to
worry
about
updating,
gitlab
and
customers.
A
It,
okay,
okay!
So
it's
in
another
third
repo
that
that's
great!
That
actually
adds
to
my
one
of
the
one
of
the
few
questions
I
have
in
the
top,
which
is
do
we
have
api
docs
for
those
it
looks
like
when
people
want
to
write,
hey,
want
to
write
the
the
tests
and
new
page
objects.
We
would
want
again
a
doc
api
docs
from
ken
lab,
and
we
just
do
it
side
by
side
and
you'd
have
to
dive
in
and
how
are
we
gonna
test
this?
B
B
Creating
method
stubs
because
ruby
can
be
very,
I
guess
the
best
word
for
it
is
magical
like
sometimes
you
don't
know
where
things
are
coming
from
and
one
way
to
get
around
this
sort
of
thing
is
stubbing,
and
so
you
see
those
with
a
lot
with.
I
think
our
spec
does
it
our
framework.
Does
it
where
it's
sort
of
it
defines
methods
by
stubs
but
actually
doesn't
have
the
implementation
it's
just
used
for
indexing
and
used
for
sort
of
documentation
like
you
refer
to.
So
my
intention
is
to.
B
If
you
look
at
that
issue,
v3,
we
have
a
website
just
like
docs.gitlab.net.com
that,
like
just
you
know
using
gitlab
pages
that
spits
out
the
api
docs,
so
you
can
just
page
by
page.
You
can
see
what
is
available.
What
elements
are
there
and
that's,
like
I
said,
that's
b3,
but
that
is
as
far
as
documentation
it's
going
to
be
sort
of
self-documenting
as
well.
A
Oh
jesus,
I
came
back
when
when
we
were
still
typing
in
what
about
have
you
thought
about
tests
sounds
like
we
moved
it
out
and
it
can
be
tested
in
its
own
own
project.
Now
think
of
this
as
like
the
if
the
test
framework
was
a
pyramid,
this
would
be
like
the
lower
levels
where
we're
testing
all
the
selenium
libraries
independently
and
it's
not
having
to
be
the
main
repo,
and
I
can
even
see
it's
being
used
in.
A
B
Yeah,
it
is,
it
is
no
small
feat
to
implement
chem
lab
in
sort
of
a
a
widely
distributed
way.
As
of
yet
there's
a
lot
of
goals
to
be
attained,
for
instance,
the
documentation
that
you
mentioned,
because
we
want
to
be
able
to
have
a
way
to
use
it
before
we,
you
know
broadly
distribute
it
docs
and
tests
included,
and
since
this
is
a
separate
repository,
it
is
a
goal
to
write
tests
for
it.
B
C
B
C
C
B
B
B
But
a
lot
of
a
lot
of
effort
is
going
to
have
to
be
into
that
and
that's
why
it's
not
intended
to
have
like
a
not
an
early
phase
but
as
far
as
long
term
and
I'm
talking
far
long
term
here
it's
it's
ideal
that
all
the
libraries
are
migrated.
You
know
to
the
to
the
chem
lab
library,
if
not
migrated,
then
at
least
duplicated,
and
then
eventually
we
could.
You
know,
switch
between
the
capybara
page
object
to
the
chem
lab
page
library
yeah.
So.
A
E
Yeah,
I
I
meant
far
in
the
future
as
well,
but
just
for
for
clarification
yeah
it
was.
You
basically
covered
what
I
was
asking.
Where
are
we
gonna
migrate,
the
old
tests
over
eventually
and
should
or
would
we
start
writing
new
tests
in
a
different
way
yeah,
while
the
old
tests
are
still
working
are
that
was
that
was
basically
my
question.
B
A
I'd
like
to
vocalize
point
c
now,
we
would
like
more
feedback
from
the
team
before
we
make
the
call.
This
is
a
balance
we
have
to
make,
because
I
see
trends
that
we
don't
want
to
leave
a
fork
in
methodology
for
too
long,
because
that
creates
more
drifts
apart
in
the
architecture
we're
doing
this,
because
we
want
to
provide
a
space
for
for
self-disruption
and
productivity
per
se,
because
we
have
a
new
opportunity
to
do
it
in
customers
dot.
A
We
find
that
it's
performant
now
we're
bringing
the
the
benefits
back
and
and
see
where
we
can
go
from
there.
So
the
folks
that
joined
the
ama
today,
I
believe,
you're
very
interested
on
on
driving
this
and
using
this.
So
I
think
we
should
do
another
gut
check
within
a
month
or
so
and
see
where
we're
at
in
the
meantime,
continue
to
provide
feedback.
A
Chem
lab
is
open
to
everyone,
correct
dan,
the
in
the
team
yeah.
I
can
foresee
that
adding
changes
in
chem
lab
and
wanting
to
test
it
out
would
be
a
two-step
process.
Now
and
that's
that's
the
the
the
drop
in
productivity.
So
if,
if
there
any
tooling
in
chem
lab,
like
a
review
app
that
you
can
test
out
your
your
new
click
or
selenium
wrappers
easily
and
then
bring
it
in
in
the
main
repo
that
would
help
there's
still
a
lot
more
questions.
E
A
B
I
want
to
have:
I
want
it
to
be
stable,
right,
docs
and
things
to
to
be
solidified
and
testing
to
be
solidified
in
chem
lab
before
a
large
rollout
and
again
as
far
as
the
rollout's
concerned,
we're
not
required
to
adhere
to
chem
lab.
We
can
do
our
work,
just
as
we
do
today
without
any
sort
of
impact
at
all
the
v1
of
what
I'm
doing
right
now.
B
You
I
have
to
in
that
specification,
go
to
gitlab
back
to
about.gitlab.com
or
staging
rather
and
sort
of
seamlessly
go
between
them,
so
the
first
spec
using
chem
lab
is
just
going
to
be
that
free
trial
for
v1
and
again
there's
no
impact
on
anyone's
work,
but
as
far
as
large
rollouts
concerned,
I
definitely
want
to
have
documentation
api
docs,
everything's
squared
away
before
you
know.
I
want
to
it's
a
big
effort,
so
I
want
it
to
be
an
actual
actual
effort
not
haphazardly
put
together
and
hacked
away
on
so
yeah.
C
B
B
E
A
A
The
design
team
had
already
had
something
for
our
ui
and
there's
a
hard
correlation
between
the
design
system
and
also
our
common
best
practices,
because
chem
lab
needs
to
work
with
pajamas,
and
vice
versa
and
pajamas
has
a
really
good
list
of
documentation.
You
can
you
can
click
on
the
the
website?
There's
a
all
the
components
are
listed.
A
B
Well,
as
far
as
speed
is
concerned,
I'm
not
sure
speed
is
the
factor
here,
because
we
we
do
have
long
running
tests,
but
you
can't
really
peg
that
on
capybara,
I
think
what
you
get
here
is
visit
our
usability
and
testability
as
far
as
water
is
concerned.
So
if
you
were
a
developer
wanting
to
use
our
our
stuff,
it's
a
lot
more
of
a.
B
B
E
A
If
I
give
it
to,
for
example,
a
new
person
that's
joining
soon,
if
writing
it
in
the
ken
lab
style
is
half
the
time,
then
digging
through
the
current
cavity
bar,
I
think
that's
a
win
and
we
can
use
that
data
to
make
the
proposal
solidify
this
for
the
changes
I
think
desiree
has
the
next
question.
Thank
you.
E
Yes,
I
was
just
wondering
about
the
qa
selectors.
I
I'm
from
what
I
asked
to
understand.
I
think
it
just
uses
the
same
thing.
Copyright
does
where
it
could
use
a
qa,
selector
or
qa
class,
because
I
noticed
we
do
still
have
some
qa
classes
that
haven't
been
updated
into
qi.
Selectors
just
want
to
clarify
that.
B
Yeah,
so
it
does
use
the
qa
selectors,
just
like
our
current
tests.
Do
the
qa
classes
instead
of
qa
selectors,
yes,
you're,
absolutely
right!
This
would
need
to
be
updated
to
to
to
go
over
and
I
think
that's
actually
a
good
design.
So
we're
not
relying
on
a
deprecated.
You
know
because
we
do
consider
the
qa
classes
deprecated.
So
it
makes
no
sense
for
us
to
sort
of
resume
a
new
piece
of
software
that
uses
that
legacy
technology
or
you
know,
legacy
selectors,
but
yeah.
B
They
would
all
need
to
be
updated
to
data
qa
selectors.
Another
thing
with
that
desiree
thank
you
for
bringing
this
up
is,
like
I
had
mentioned
with
the
page.
Capybara
page
object
versus
camera,
page
library-
if
you
recall
from
me,
sharing
my
screen
and
I'll,
do
it
just
again
just
to
just
to
make
this
clearer.
B
This
is
the
copy
bar
page
object.
We
have
that
suffix
on
field
right,
so
first
name
field,
whereas
in
here
we
have
first
name
as
a
text
field
and
it's
omitting
the
field.
This
could
present
a
problem
which
we
might
have
to
update.
I
haven't
really
thought
this
far
ahead,
but
we
might
have
to
update
those
existing
data,
qa
selectors,
to
remove
the
suffix,
which
is
fine
really
in
my
opinion
anyway,
because
those
it's
the
more
you
pair,
your
selectors
with
the
element
type
the
worse
off.
B
You
are
because
you
get
into
sort
of
the
territory
of
xpath,
where
you're
sort
of
looking
for
specific,
very
specific
things,
which
is
fine,
but
you
also
want
to
leave
it
vague
enough
in
or
so
you.
So,
if
you
update
a
div
say
to
a
span,
you
want
to
be
able
to
account
for
that,
but
I
don't
know
I.
I
have
to
think
that
far
ahead,
I
haven't
yet,
but
that
isn't
all
that.
That
is
also
another
thing
that
that
I
need
to
consider
with
the
data
qa.
A
A
I
can
see
the
case
being
both
because
each
framework
has
its
own
business
case
on
what
what
to
test,
and
it
should
be
agnostic
of
how
how
the
browser
connects,
but
I
do
see
a
case
of
having
handling
the
different
types
of
browsers
in
canada,
because
that's
where
most
of
the
challenges
are
you
test
on
ie
10,
for
example,
the
click
method,
the
file
element
may
be
different
per
se.
So
what
are
your
thoughts
on
on
how
this
will
fit
with
browser
cross
browser
testing.
B
Yeah,
I
think
it's
important
to
distinguish
that.
Chem
lab
is
not
a
test
runner
and
it
is
not
a
water
replacement
or
anything
like
that.
It's
not
designed
to
be
sort
of
a
another,
at
least
at
first,
an
alternative
to
say
water
or
capybara.
It's
not
it's
not
like
it's
a
its
own
thing.
What
it
is
is
a.
B
So
as
far
as
cross
browser
compatibility
is
concerned,
if
you
look
well,
I
mean
I
haven't
linked
the
mr
yet,
but
I
had
mentioned
earlier
that
the
chem
lab
page
libraries
just
reuse,
the
the
web
driver
that
capybara
is
using
because
of
that
right.
Now
it's
just
piggybacking
off
of
what
gitlab
decides
to
do
with
it.
In
gitlab
we
already
have
the
ability
to
run
cross
browser.
We
don't
do
it
much,
but
we
do
have
that
kind
of
capability.
B
A
Okay,
so
what
I'm
hearing
is
the
the
repo
under
test
is
called
that
way:
customers.
or
gitlab
the
main
repo,
the
call
to
connect
the
runner,
the
browser
runners
is
that
specific
to
that
repo,
and
we
will
leave,
leave
it
there
for
now
that
they're
in
charge
of
doing
it
the
job.
This
is
just
a
more
productive
design
pattern.
B
And,
like
I
said
at
least
for
v1,
it's
not
my
intention
to
sort
of
create
an
a
widely
used,
I'd
like
it
to
be
widely
used,
but
it
because
I
want
git
lab
to
be
sort
of
a
forefront
as
a
as
a
template
for
other
companies
to
use,
rather
than
the
other
way
around.
I
don't
like.
I
don't
like
in
principle
us
looking
at
other
companies
to
say
hey
what
are
they
doing?
Let's
replicate
it.
No,
we
need
to
implement
and
iterate
and
and
greenfield-
and
you
know,
build,
not
buy.
That's.
B
A
Mike
don't
treat
my
comment
as
a
prescription,
I'm
merely
asking.
I
do
think
that
having
the
repo
under
tests
define
their
cross
browser
coverage
charter
is
is
better
at
this
point.
We
have
to
deal
with
the
how
how
it
would
look
like
when
we
bring
out
the
knowledge
of
hey.
I
know
that,
if
you're
on
edge,
it
has
to
call
fine
element
two
times
with
like
a
another
sleep
in
the
middle
versus
one
time
in
in
chrome,
that
sort
of
thing
that
could
be
a
technical
fallout.
C
Yeah
so,
for
example,
I
wrote
a
test
yesterday,
but
I'm
gonna
have
a
few
more
coming
up
shortly
and
I
just
I'm
curious:
do
we
have
do?
We
have
license
to
go
ahead
and
start
writing
tests
in
chem
lab
against
git
lab
or
is
it
are
we
still
considering
this
to
be
experimental
and
then,
for
example,
if
I'm
bringing
the
runner
team
on
board
I'd
love
to
just
point
them
to
one
framework,
as
opposed
to
two.
A
Well,
well,
that
is
fine.
The
link
open
for
contribution.
I
think
the
good
step
forward
is
our
own
validation,
that
this
is
more
performant.
No
sorry
more
productive,
my
performance,
more
productive
for
our
team.
For
for
the
other
repos,
where
we
need
a
selenium
connection
and
a
design
pattern
it
could
it
could.
We
could
reach
the
decision
faster
or
slower,
depending
on
how
he
goes
from
here,
but
I'm
up
for
us
experimenting
and
using
it
in
other
places,.
B
B
Because
again,
this
is
the
this
is
the
first
iteration.
Then
it
will
be
usable,
and
so
there
are
a
lot
of
things
to
be
considered.
For
instance,
I
want
documentation.
I
wanted
to
be
clear.
You
know
how
to
use
it
before
it's
widely
distributed
and
so
to
your
question
of
using
it.
Do
you
have
license
to
use
it?
The
problem
is
with
a
sort
of
a
starting
green
field
like
this.
B
B
B
B
B
Okay,
someone
help
me
like
y
combinator.
You
have
people
that
sort
of
challenge
the
status
quo
of
how
things
are
done.
What's
that
called
rubber?
It's
it's
disruptors.
Yes,
thank
you.
It's
very
disruptive
because
it
it
goes
against,
like
the
the
zeitgeist
of
of
how
page
libraries
and
page
objects
are
written
and
understood,
and
again
one
of
the
downfalls
that
I
think
of
page
object
design
is
that
it
requires
specific
elements
for
sorry,
specific
methods
to
form
specific
actions
against
those
methods
against
those
elements.
A
If
I
were
to
recap
this,
the
page
object
is
still
the
duty
of
the
repo
under
test,
how
you
defined
those
classes,
but
now
the
methods
inside
those
page
objects
is
a
library
now
chem
lab
so
you're,
providing
those
those
mechanisms
that
we
can
build
page
objects
faster
where
faster.
Hence
the
library
is
that
roughly,
where,
where
are
you
going.
B
A
B
That's
that
design's
fine,
but
the
problem
that
you
commonly
run
into
is
that,
if
is
that
those
objects
will
have
to
be
interacted
on
upon
somehow
and
usually
those
elements
are
interacted
upon
by
modifying
the
object
in
methods
so
in
ruby
set
first
name,
and
then
you
provide
an
argument
for
the
first
name
and
then
that
calls
the
elements
that
cause
the
object
to
fill
the
element.
That's
a
huge
layer
of
abstraction,
that's
unnecessary
and
it
is
it's
redundant.
B
It's
it's!
It's
inefficient.
So
the
page
library
design
pattern
is
more
of
you
define
your
elements
similar
to
how
you
do
with
page
object.
But
you
you,
don't
you
you
don't
worry
about
the
the
sort
of
redundancies
that
you'd
find
in
the
page.
Object,
design
and
again
it's
it's
hard
for
me
to.
I
do
have
a
documentation,
a
document
written
for
it
like
a
drafted,
but
does
that
sort
of
make
sense
like
if
you
have
the
page
object
and
you
have
the
page
library.
A
A
Having
adapter
into
into
that
terminology,
helps
think
of
that
think
of
how,
for
example,
if
if
we
use
this
in
interviews,
people
were
not
familiar
with
scratching
their
heads.
What
is
page
libraries
I'm
using
page
optics
for
for
a
really
long
time,
so
please
be
mindful
of
that,
as
we
introduce
innovative
ideas
and
disruption.
B
Yeah,
I
guess
yeah
in
in
my
view,
it's
the
page
object.
I
will
keep
that
in
mind.
The
page
object
has
been
sort
of
an
ambiguous
way
of
describing
how
to
layout
your
elements
in
a
page
in
selenium,
particularly,
but
the
problem
is,
is
that's
so
ambiguous.
Some
people
have
specific
types
of
specific
ways
of
writing.
Page
objects,
but
there
is
an
ultimate
because
it
is
a
design
pattern.
B
C
Yeah,
I
had
one
more
question:
I'm
sorry,
I
buried
it
in
a
sub
point.
I
I
apologize
under
dan's
last
comment,
so
so
we're
using
the
the
terminology
page
libraries
is
there
I'm
not
finding
much
or
really
anything
about
page
libraries
as
a
design
paradigm
is
this?
Is
this
page
factory?
Is
this
related
to
page
factory
by
chance
or.
B
You
will
not
find
anything
referring
to
page
libraries
online
this.
So
when
I
started
designing
my
own
page
libraries,
I
made
a
distinction
between
how
the
pattern
was
starting
to
emerge.
The
page
object
pattern
and
the
way
that
I
did
things-
and
I
felt
I
I
noticed
immediately-
the
downfalls
and
the
pitfalls
of
the
page
object
as
far
as
maintenance
is
concerned,
because
I'm
sure
all
of
us
have
dealt
with
page
objects
being
very
hard
to
maintain
updating
methods
and
renaming
refactoring.
B
All
this
stuff
is
very
maintenance
driven,
and
it's
not
a
very
good
design.
If
you
can't
maintain
it,
and
so
page
library
is
a
design
that
I
had
that
I
had
concocted,
and
I
got
the
term
library
because
that's
in
computer
science,
that
is
the
term
for
for
a
bucket
of
code.
That
does
something
specific.
So
you
have
a
library
of
this
library
of
that
and
that's
the
that's
the
computer
science
terminology
for
it,
and
so
it
was
a
design
pattern
that
I
never
actually
you
know
got
to.
B
B
Yeah
for
sure
I
also
intend
to
create
a
a
diagram,
not
a
flowchart,
but
a
diagram
to
to
show
how
the
design
will
work
in
get
lab.
Specifically,
as
far
as
those
my
that
modularized
approach
that
I
had
mentioned
sort
of
how
those
things
are
laid
out
and
yeah,
I
noticed
desiree,
you
had
a
question:
do
you
want
to
go.
B
Great
great
question,
and
that's
actually
one
of
the
things
that
I've
been
sort
of
considering
how
to
fix
this.
That
is
a
gap
that
this
would
introduce
is
like.
If
we
don't
specify
the
view,
let
me
back
up
and
say
the
advantage
of
the
view
of
how
we
do
things
now
is.
If
a
developer
changes
a
front
end,
it
will
be
caught
immediately
if,
if
our
test
refers
to
that
object
or
that
you
know
that
element
and
it
has
changed
or
it
is
not
there
anymore-
that's
a
really
good
tie-in
with
the
source.
B
So,
as
it's
implemented
right
now,
can
we
still
use
the
qa
selector
test?
The
answer
is
no,
as
it
stands
right
now,
but
I,
but
I
intend
to
sort
of
brainstorm
on
an
agnostic
approach
to
that,
where
not
just
gitlab,
but
customer
stock
could
use
that
sort
of
pattern
to
to
to
to
check
parity
between
their
the
front
end
and
the
tests
in
the
page
page.