►
From YouTube: YUI Open Roundtable 12/20/2012
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
C
C
So
we
have
testing
that
we're
doing
for
the
testing
we're
doing
with
Getty.
We
have
a
lot
of
work
that
we
did
today
with
the
that
I'm
doing
with
Dave
on
munching
browsers
and
I
was
going
to
show
some
of
that.
But
I
can't
steer
my
screen
for
some
reason
so,
but
yeah
we're
able
to
run
lots
of
lots
of
tests
using
browsers
that
start
and
stop
after
the
test
run,
which
is
really
great
because
that's
a
big
concern
for
us
is
having
for
Percy
I
purposes
like
automating,
the
starting
and
stopping
up
browsers.
C
C
Yeah,
so
we
we
had
like
Dave,
helped
me
out
a
lot
with
setting
up
selenium
on
some
post
that
we
have
internally
for
and
so
we're
you
know
this
is
this:
we
wanted
to
try
using
selenium
as
a
browser
launcher
just
so
that
you
know,
because
it's
reliable
not
only
that,
but
we
wanted
to
explore
what
we
could
do
with
sauce
now
that
they
offer
free
testing.
So
we
should
just
be
able
to
change
the
host
to
sauce
and
be
able
to
run
tests
using
Yeti
on
their
infrastructure
as
well.
So
yeah.
D
C
F
A
So
I
contacted
Marsha
Permian
and
not
get
response
and
he's
in
Brazil
I've
imagined
that
this
is
for
the
the
next
topic,
which
is
no
gap,
native
apps
seems
likely
something
that
we
will
spin
off
as
a
separate,
hang
out
I'm
going
to
find
out
from
him
what
is
a
good
time
to
date,
for
you
two
to
connect
up
with
us
than
that
so
going
to
table
the
phonegap
I
hadn't.
For
now
the.
A
We
want
to
talk
about
yeah,
the
first
to
shove
it
over
the
forward
mess.
I
remember
there
was
a
question
about
that,
so
me
offers
so.
There
was
the
we
have
linen
suit
here.
Who
gave
a
talk
from
Y
to
accomplish
when
we
talk
and
regarding
yacht
and
as
a
result
of
posting
the
video
up
on
the
blog,
we
got
some
questions
found
ya,
somos
una
lets.
You
have
chance
to
chat
sure
I.
F
Don't
think
that
it's
basically
the
Eric
Pearson,
he
posted
a
comment
on
the
other
stuff
and
the
one
he
looks
like
he's
dangerous
to
be
that
JSON
schema
part
because
you're
thinking
him,
we
said
to
auto-generate
models
and
getting
stuff
for
his
application.
So
I
guess
the
question
is
that
I
got
a
wider,
thoughtful
talk
to
them.
Warranty,
ostensible
reason,
use
cases
and
McCarthy
details
so
not
sure
he
is
selected.
Yes,
come
I!
Guess
how
can
we
connect
they
good
contact
with
him?
If
I
want
more
details.
F
F
So
so
we
didn't
discuss
the
starbucks
him
a
lot
of
stuff.
First,
the
schema
we
originally
we
meet,
we
shall
have
schema,
was
to
use
it
as
a
way
of
defining
what
in
a
motorbike
going
for
finance
portfolio
site
as
a
lot
of
competitive
am
homeless
right.
So
it's
they
better,
not
read
it
down
into
that
JSON
schema.
It
helps
us
with
sharing
that
with
within
the
team,
or
you
know
with
excellent
teams.
F
So
that's
so
that
those
what
what
that
originally,
why
we
had
in
there
and
then
you
can
should
evolve
into
one
at
that.
But,
for
example,
we
are
using
that
we
have
stated
it
tells
it
can
use
the
honor
generate
the
the
attributes
for
a
model
so
in
the
education
actually
have
pleasure.
Identify
dinners.
Origin
is,
if
you
say
the
schema
mcknight,
that
the
skill
is
a
number
and
arranged
suppose
depending
upon
21.
So
we
could
isn't
it
during
a
reporter
for
that
for
the
attribute
in
a
bottle
later,
potentially
we
can
also
use
them.
F
Is
this
q
on
to
check
against
this
object,
see?
What
is
what
is
that
it?
Whether
they
don't
know,
though,
we
are
poor
or
just
she's,
that
pretend
identity
that
for
generating
a
ventilator
to
buy
the
new
attribute
value,
whether
you're
trying
to
change
the
value
of
the
model.
So
that's
that
Harmon
officials
asking
whether
can
break
this
off
into
your
pieces
and
he
can
probably
use
for
his
vision.
F
F
C
Yeah,
so
just
so,
I'm
understanding
this
right,
the
JSON
schema
part
essentially
allows
a
developer
to
describe
the
attributes
and
potentially
relationship
between
it's
in
what
type
of
data
is
expected
image
in
a
JSON
format,
which
then
there's
some
there's
some
code
that
will
then
process
that
specific
JSON
format
and
create
just
the
attributes,
object
out
of
it
or
create
like
the
model
subclass.
F
C
C
J
J
I
It
or
yeah
yeah,
so
that's
I,
think
we
probably
it
chocolate
every
more
go
through
the
design
and.
A
F
F
C
The
way
so
I
took
a
similar
path
with
a
couple
of
the
sink
layers
that
we
have
for
model
and
model
list.
I
put
them
in
the
gallery.
First
got
a
bunch
of
feedback
on
them
for
six
months
or
something
and
then
decided
that
we
should
move
the
rest
one
into
core
just
because
it's
so
common
and
it's
nice
to
provide
it
out
of
the
box
and
then
I
had
a
chance
to
essentially
act
on
all
the
feedback.
C
I
got
from
those
six
months
of
people
using
it
to
make
it
even
better
for
the
version
that
actually
went
into
the
core
gave
me
a
chance
to
like
refactor
the
code
and
add
any
other
features
that
didn't
exist
like
I
had
I
added
I
believe
the
cross-site
request.
Forgery
features
baked
right
in
so
people
can
protect
against
that
standard
pack.
That
or
you
know,
exposure
that
people
have.
C
But
I
think
that
when
looking
at
it
for
the
gallery,
you
could
write
it
almost
like
that
it
exists
as
part
of
Yui
or
like
hanging
it
off
of
the
model
name
space,
and
I
can
show
you
examples
of
us
to
doing
that.
Where
we
have
these
model
extensions
that
hang
off
of
wideout
model
and
that,
but
then
still
that
way
like
in
miata,
you
can
start
you
guys
can
start
consuming
that
gallery.
A
F
A
A
Yeah
once
you
once
you
get
something
like
some
better
check,
anything
like
that
will
be
able
to
the
repercussions
from
then
I
don't
see
anything
from
IRC,
so
we
can
move
on
before
I
jump
into
pull
requests.
I
was
going
to
check
the
sea
just
before.
If
there
were
other
read
brave,
to
give
you
demos
working
our
my
check
anything
else
at
Before,
we
jump
into
them
a
little
request.
C
Yeah
think
so,
but
yeah
can
I
just
do
mine
after
you
get
that
stuff
done
sure
cool.
So.
A
A
What
happened
with
these
four
requests?
I
said
the
email
to
the
contributors,
mailing
list
or
feedback
for
hand
and
got
could
have
been
a
good
feedback
in
terms
of
they'd
mentioned
that
two
of
these
are
pretty
much
ready
to
go.
They
were
approved
by
the
owner
while
back
and
they've
got
past
the
72-hour
limitation
already.
So,
with
that
this
for
quest,
139
and
I
want
you
to
do
is
basically
find
an
owner
who
will
join
in
it's.
A
D
A
D
C
Looks
like
some
of
the
logic
in
here
is
different
than
just
exposing
the
method
on
the
y
namespace,
and
it's
also
on
why
like
dispatch,
iterator
I,
don't
know
I
mean
we,
we
should
probably
figure
out.
If
that's
the
thing,
that's
the
name.
We
want
to
support
forever,
but
okay,
like
it
makes
sense
for
us
to
expose
this
from
what
I
gather
that
it
does
but
dispatch
iterator
sounds
weird.
Do.
C
This
is
the
whole
thing
that
we
have
in
here,
where
we
encourage
people
not
to
use
Y
dot
each
but
to
use
wide
array
dot
each
or
why
dad
object
at
each,
and
so
this
is
exposing
essentially
the
guts
of
how
we
determine
the
type
and
then
based
on
their
call
into
the
particular
method.
So
I
don't
know
you.
C
C
J
J
J
All
right
I
need
it
for
my
functional
programming
gallery
module.
It
would
be
nicer
if,
if
you
use
it
for
my
body
and
then
yeah
I
me
back
in
January,
Ryan
did
say
I
feel
kind
of
weird
about
this
since
I,
really
don't
like
dispatch
I've
actually
been
hoping
to
get
rid
of
it.
It's
only
used
by
why
not
eating
white
out
some
and
I
think
we
should
discourage
the
use
of
some
of
those
methods
and
instead
encourage
people
to
use
the
appropriate
methods
on
whatever
your
white
object
and
then.
J
J
A
J
So
I
think
that
the
objection
is
actually
in
the
common
history,
so
we
can
treat
that
as
an
objection
that
still
needs
to
be
resolved,
so
we're
not
okay
to
merge
acento.
Basically,
I
think
it's
on
either
John
shoulders
or
someone
else
hoop
that
supports
merging
this
in,
to
convince
us
that
that
there's
value
to
exposing
this
particular
API,
rather
than
using
the
the
ones
that
already
exists,
I
want
there's.
E
E
E
B
J
You
could
close
that
reference,
the
conversation
in
this
in
this
book
and
and
add
your
comment
to
this
of
history.
Then
maybe
he
could
try
in
this
way
and
then
come
up
with
something
we
can
all
be
happy
with
for.
A
A
We
agreed
on
the
action
visible
you
thanks
to
them:
okay,
the
next
one
that
was
listed
as
good
to
go,
but
maybe
we
should
have
more
attention
to
was
the
quest
48
from
run
back.
That
says,
it's
a
fixed
forward:
ticket
25,
28,
7
30
to
40
minutes
description
is
attributed,
says,
default,
values
that
are
returned
invalid
value
and
it
would
be
initialized.
F
G
C
C
J
J
C
C
So
yeah,
oh
yeah,
there's
hope
actually.
C
This
seems
like
definitely
something
we
should
do,
but
we
may
have
to
I,
don't
think
attribute.
J/S
has
any
code
in
it
anymore,
yeah.
It
doesn't
have
any
of
this
code.
So
all
that
codes
an
attribute
core.
L
J
J
A
J
B
J
J
B
J
B
L
C
C
So
just
Sutter
yeah,
so
I
think
that
this
audience
comment
was
not
addressed
by
this,
because
if
this
failed
allegation
it
doesn't
get
reset
back
to
the
right
value.
This
is
only
if
the
center
has
invalid
value
returned,
so.
C
E
You
file
it
that
come
in.
What's.
B
A
A
J
A
A
B
G
J
E
C
E
C
Yeah,
so
this
is
something
yeah.
We
should
just
take
the
the
spirit
of
this
change
and
possibly
the
test
too.
But
the
main
thing
I'll
want
to
double
check
on.
Is
the
coverage
of
this
code
path
to
make
sure
that
we
have
a
hundred
percent
coverage
in
this
area
for
this
change,
but
it
looks
like
we
could
yeah
use
this
test
so.
C
J
B
A
J
A
A
Here
we
go
sticky,
okay.
Well,
let's
see
we
have
anything
on
our
see
channel,
welcome
questions,
no
okay!
Well,
I!
Don't
think
everybody
for
coming,
sir,
so
you
talked
about
yeah
and
I.
Share
wants
feedback,
especially
the
great
comments
on
the
floor
request
in
terms
of
next
week.
Do
we
know
it
will
be
hanging
around
here.
C
About
I
show
this
real,
quick,
all
right,
cool
yeah.
Now
that
my
screen
sharing
is
working
so
yeah
yeah,
oh
man,
is
it
not
gonna
work
now
seriously?
Oh
there,
it
is
okay,
cool
yeah.
E
C
So
this
is
what
Dave
and
I
have
been
doing
today,
as
we
are
wanting
to
have
our
test
run
as
quickly
as
possible,
and
so
we
have
right
now
is
we
have
browser
launching
happening
with
a
8
posts
right
now
that
have
chrome,
firefox
and
ie,
and
so
we
tested
out
with
with
actually
all
a
kind
of
all
day
to
day
we've
been
running
Yeti
on
all
these
different
hosts
that
we
have
and
yeah
it's
we've
got
right
now,
it's
running
as
fast
as
running
our
6c
6330,
two
tests
that
run
in
three
minutes
and
37
seconds
on
Chrome.
C
So
that's
how
fast
we
want
our
are
to
be
internally
for
just
just
on
Chrome
with
other
browsers.
That's
what
we're
looking
at
right
now
is
see
how
fast
we
can
get
with
with
other
browsers
running
them
all
in
parallel,
so
I
could
show
what
we
have
when
this
finishes.
So
three
minutes
37
seconds
for
6000
that
rat
there
622
tests
so
yeah.
This
is
happening
so
fast
that
we
actually
yeah
can't
like
it
just
causes
this
whole
screen
to
go
crazy.
It's
lots
of
our
DN
host.
C
We're
hoping
to
get
this
like
use
stuff
in
Travis
to
like
so
I'll
show
what
this,
what
this
is
doing,
because
just
seeing
that
this
running
in
our
CI
system,
isn't
it's
fun
so
right
now,
what
this
is
doing
is
it's
actually
launching
eight
Chrome's
Yeti
is
actually
launching
eight
Chrome's.
This
is
available
right
now
on
my
fork
of
Yeti
and
now
it's
started,
and
this
is
running
so
fast
that
I
could
barely
even
see
ya
every
time.
There's
an
update.
That
is
a
test
passing
or
failing
right
now
charged
right.
C
So
this
is
the
slowest
that
it
gets
is
as
at
registered
charts.
Thank
you
sure
yeah,
but
it's
yeah
with
charts.
It's
taking
three
minutes
36
seconds
yeah.
So
this
is
a.
This
is
right
now
running
on
our
internal
build
slave,
and
so
we're
going
to
have
the
next
step
obviously
is
to
get
this
running
on
Travis
too.
C
So,
what's
really
neat
about
this
is
that
we
were
using
Yeti
before,
but
we
were
only
doing
it
on
demand,
and
so
now,
what
we're
going
to
be
able
to
do
is
have
this
replace
our
existing
selenium
based
browser
launcher
that
also
runs
tests
and
replace
the
thing
that
rides
the
test
with
Yeti
and
so
we're
launching
with
selenium
the
browser.
But
all
the
tests
that
are
happening
inside
are
happening
with
Yeti.
C
That
turns
out
to
be
really
great
because
we're
finding
you
know
by
using
this
or
in
parallel
and
not
having
to
write
any
tedious
like
Java
code
and
said
this
is
all
happening
with
by
using
the
web
driver,
Jason
API
and
then
where,
instead
of
having
to
do
all
tedious
bridge
stuff,
we're
just
doing
that
all
inside
of
Yeti
with
for
collecting
test
results,
yeah,
we
just
use,
probably
the
thing
that
selenium
is
best
used
for
which
is
automating
a
browser
to
open
and
close
it
and
now
we're
doing
the
rest
all
with
Yeti.
C
So
yes,
this
is
about
halfway
done
now.
Two
minutes.
C
C
A
bug
in
and
red
hat
enterprise
linux
is
preventing
the
yeah
it's
preventing
the
feedback
line
here
from
clearing.
So
it's
every
update
is
not
getting
cleared,
but
if
I
ran
this
with
a
J
unit
yeah,
you
would
just
see
all
the
XML
come
through.
Instead
of
this,
this
mess
right
here,
which
is
the
which
is
pretty
pretty
cool-looking.
We've
been
staring
at
this
kind
of
all
day,
watching
our
fastest
gets
so
so
yeah.
We
have
all
these
Chrome's
in
use.
The
next
step
is
ya
de.
C
Have
do
this
all
in
Firefox
that
you've
already
done
and
that
takes
a
little
bit
longer
about
seven
minutes
or
so
so
yeah,
let's
see
how
much
longer
is
this
gonna
take
a
hint
it's
like
going
so
fast,
35
seconds,
so
yeah
just.
G
C
Version
version
numbers
for
chrome,
since
they
you
know
change
so
frequently.
Yeah.
Are
you
guys
gonna
have
some
scheme
to
say,
like
chrome,
greater
than
you
know,
or
just
do
like
the
major
version
number
for
chrome
or
something
and
just
consider
a
wash
for
all
the
other
one
so
that
it
can
split
up
the
test
amongst
Chrome
browsers
that
are,
you
know
very
by
like
one
or
two
days
from
their
update.
Yeah
Dave
brought
that
up
actually
yesterday
to
about
how
there's
a
I
think
iOS
6,
0,
2
or
6
yeah.
B
C
12,
if
there's
a
there's
a
minor
release,
that's
only
on
some.
You
know
only
like
a
certain
model
has
and
so
yeah
the
awesome
to
for
Yeti
to
understand
what
is
similar
enough
and
so
actually
I.
Think
Dave
had
a
really
good
point
is
just
instead
of
doing
what
yet
he
does
now
is
just
haven't
breaking
down
on
the
capabilities
of
each
browser,
that's
connected
to
Yeti,
and
then
you
could
specify,
and
yet
his
config
file
that
you
want
these.
You
want.
C
You
want
this
major
version
or
whatever
you
know
you
just
want
chrome,
any
chrome
version
or
you
know,
chrome,
that's
you
know
with
December
thing
or
what
have
you
so
so
this
town
took
three
minutes
14
seconds
now,
which
is
pretty
cool.
I'll
share
my
screen
again,
so
you
can
see
what
I'm
looking
at
but
yeah
it's
sometimes.
C
I
think
that
the
Yeti
crashed,
but
it's
actually
it
just
finished
so
fast
that
I
can't
catch
it
when
trying
to
look
for
it,
but
yeah
3,
minutes,
14
seconds
RW,
so
yeah
well,
91%
lion
cover
each
other.
Yeah
I
wish
I
wish
yeah,
but
that
was
failures.
We
had
one
failure
of
all
and
all
those
so
yeah,
but
you
fail.
That's
a
good
question.
Yeah
I'm
gonna
withhold
that
no
I
just
don't
know
so.
G
G
C
Yes,
yeah
so
yeah,
so
that's
that's
it
like
we're
like.
Let's,
let's
let
me
do
this
one
more
time,
so
I
can
show
what
this
looks
like
with.
I
can
get
my
windows
to
focus
properly.
Here
we
go
yes.
This
is
it's
kind
of
hard
to
show
all
this
happening,
but
yeah,
there's
eight
here
so
you're
gonna
go
over
here
and
start
start.
This
again,
a
real.
A
Question
from
IRC
they're
asking
it
you're
saying
the
test
on
chrome
went
from
six
minutes
for
three
minutes.
You.
A
C
I'm
saying
is
that,
when
we're,
when
we're
running
with
this
many
browsers
at
the
same
time,
what's
happening,
is
that
Yeti
is
spreading
out
the
load
of
of
the
project
like
all
the
different
HTML
files
we
have
240
of
them.
I
believe
around
that
much
and
it's
actually
all
eight
of
these
are
running
different
tests,
so
they're
running
eight
kind
of
eight
tests
to
run
concurrently
at
a
time
and
the
reason
yeah
anyway,
so
we're
running
eight
of
them.
C
So
it's
you
know
cuz,
8,
16
and
so
forth
until
it
gets
to
the
end
until
there's
nowhere
to
run,
and
so
what's
with
with
running
that
mini
at
the
same
time,
we
can
finish
our
testing
in
about
three
and
a
half
minutes.
So
so
that's
what
this
is
showing
is
yeah
that
the
ETA
is
three
and
a
half
minutes,
but
it's
going
to
finish
faster
than
that.
C
G
C
So
this
is
on
a
very
txt
machine
already,
that's
that's
we're
sharing
this
host
with
with
a
lot
of
other
people
and
so
yeah.
It
can
be
a
lot
faster
than
this.
We're
running
these
on
unreal
hardware,
but
this
I
mean
this
shows
what
we
can
do
if
we,
if,
like
obviously,
we
can't
get
our
Travis
builds
this
fast
with
what
people
can
see
with
that
with
what
we
can,
but
not
with
what
sauce
provides
right,
because
sauce
only
gives
us
three
and
in
parallel
right
we
need
on.