►
From YouTube: YUI Open RoundTable 03/21/2013
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
A
See,
if
welcome
to
why
up
and
around
table
you're,
watching
this
DNA
of
e
or
layer
in
the
future,
we're
all
know.
A
D
C
So
I
was
talking
to
Eric
about
this
about
a
week
or
two
ago,
I've
been
working
on
these
modules
in
Harris's
back
to
me
and
gave
me
some
good
feedback,
so
I've,
just
iterated
bit
and
I
want
to
just
talk
a
bit
about
it.
It
actually
got
pushed
in
the
lap
in
the
latest
gallery,
push
which
was
yesterday,
but
I've
done
some
work
since
then.
C
Actually
so,
anyways
I've
been
making
I
actually
made
three
gallery
modules:
a
widget
extension
called
y
dot,
widget
pointer
and
then
a
tooltip
that
sits
on
top
of
that
which
is
a
wide
I
which
it
a
subclass
and
then
a
wide
I'll
pop
over
which
extends
wideout
tooltip.
So
I'm
just
going
to
kind
of
made
this
little
demo
page
up
and
I
just
wanted
to
go
through
it,
and
this
is
more
to
just
kind
of
show.
C
You
guys
what
I've
been
working
on
and
also
I
did
have
some
questions
regarding
skinning
and
just
general
our
organization
of
gallery
modules
and
also
testing
so
I'll.
Just
shout
I'll.
Just
tell
you
guys
about
that
as
I
go
through
it.
So.
C
That's
a
good
question,
so
the
problem
is
like
my
github
repo
does
have
so
I've
this
I've,
this
repo
called
tipsy,
which
has
the
gallery,
which
has
the
tooltip
and
the
pop
over
and
have
another
repo
called
widget
pointer.
That
has
the
widget
extension.
The
only
issue
is
so
these
are.
These
are
shifter
like
these
are
both
using
shifter
and
their
gallery
modules,
but
I
don't
actually
have
the
bill
directories
on
here.
That's
just
the
storage
directories.
C
So,
as
a
result
like
when
Run
ships
are
built,
it
like
wraps
the
Yui,
add
statement
around
everything:
I
don't
have
that
file
on
github,
which
is
why
this
is
pulling
it
from
my
local
gallery
build.
So
if
you
put,
if
you
pulled
this
down
and
built
it
using
shifter,
then
yes,
you
could.
This
is
just
the
manual
test
page
that
I'm
running
right
now,
which
has
like
a
bit
of
documentation
on
it
right
now.
The.
C
I
mean
it's
still:
it's
still
kind
of
half-baked
I
was
actually
gonna
like
the
full
release.
It
I'll
submit
a
CDN
request
for
the
next
gallery
push
next
week.
This
is
more
for
just
like
for
some
things
that
I
wanted
to
clear
up
and
kind
of
a
little
mini
organizational
code
review
sort
of
thing,
but
yeah
it
is
on
github.
So
you
can.
C
You
can
kind
of
pull
it
down,
it's
widget,
pointer
arm
and
tipsy,
which
is
which
that,
but
I
don't
know-
I
don't
know-
read
these
and
stuff
yeah
that
stuff
that
I
have
to
do
this
week.
I
do
have
kind
of
a
readme
here,
though,
so
it
is
on
the
gallery
but
hold
off
until
next
week.
C
I
think,
because
kind
of
I
can
added
some
stuff
today
so
essentially
what
it
is
in
is
a
tooltip
that
so
maybe
I'll
go
down
to
the
demo
part
first,
so
essentially
tooltip
that
you
can
that
you
can
place
around
your
content
above
below,
left
and
right
and
in
this
widget
pointer
extension,
adds
this
little
pointer
thingy
that
you
can
see
there
till
like
a
bounding
box.
So
you
can
actually
use
the
pointer
extension
with
any
widget.
C
C
The
pointer
has
to
point
to
the
right,
so
it
does
that
automatically
and
it
also
you
can
also
kind
of
I'm,
still
writing
up
the
docks,
but
you
can
essentially
set
the
placement
dynamically.
A
little
figure
things
out
the
most
basic
way
to
use.
It
is
just
to
go
tipsy
like
me,
while
tipsy
with
a
selector
and
Eric
made
a
good
point
when
I
talked
to
him
about
this
last,
like
two
last
week
or
three
or
four,
which
is
there's
only
ever
one
instance
of
this
thing
created.
C
So
if
you
have
like
a
hundred
blue
tags
with
rel
tipsy
there's
only
one
of
them,
that's
going
to
be
created
like
one
widget,
that's
going
to
be
created,
it's
just
going
to
keep
moving
around
between
all
of
them.
If
you,
so
you
essentially,
you
create
a
one
widget
for
a
single
selector,
so
I
have
like
these.
These
four
are
all
using
one
instance
of
tipsy,
but
then
this
one,
which
is
a
clicked
one,
is
using
a
different
version
of
it
excited.
A
C
A
C
Just
hiding
it
it's
all
it's
just
it's
just
toggling
class
names
and
the
only
thing
that
it's
doing
here
is
getting
the
getting
the
content
and
setting
that
and
I
could
probably
add
some
smarts
to
it,
to
figure
they
add
like
a
little
hash
or
something
which
can
figure
out
if
it's
already
been
there
doesn't
have
to
interact
with
the
dog,
but
I
haven't
done
that
yet
so
20.
E
And
I'm
sure
you
you
know
about
this,
your
pointer
seems
to
be
hanging
over
the
thing.
Is
it's
yes,
yes,.
C
So
I
have
to
build
in
offsets
and
this
is
tricky
because
actually
this
brings
me
to
another
point
ahead,
which
is
so
as
for
the
actual
alignment.
I'm
using
widget
position
align
to
actually
position
this
widget
in
accordance
with
this
node,
but
then
for
the
actual
pointer,
I'm
using
the
alignment
on
the
node
align
plugging
because
the
pointer
itself
is
just
a
dom
node.
So
the
pointer
is
using
on
a
line
plug
in
to
position
itself
rig
with
respect
to
the
bounding
box
of
the
widget
that
make
sense.
C
C
I
was
actually
thinking
of
doing
something
like
that
when,
when,
when
the
alignment
values
are
calculated
so
yeah,
that's
definitely
one
thing
we
do.
What
the
other
thing
I
want
to
ask
is
this
widget
position?
Align
thing
is
well.
One
of
the
things
is
the
alignment
plugin
has
this
parameter,
which
will
keep
it
aligned
when
the
browser
resizes,
but
the
widget
position
align,
think
doesn't
and
I
feel
like
they
both
to
have
that,
so
I
might
tend
to
pull
request
to
pull
that
in
yeah.
E
E
C
Their
API
is
slightly
different
to
like
the
way
do
we
use
it
and
I
feel
I
wish.
I
could
just
use
one
of
them
because
then
I
mean
I
could
just
use
nodal
line,
but
it's
just
easier
to
just
mix
in
a
positional
line,
but
because
right
now,
I'm
pulling
in
both
of
them
and
I.
Don't
really
need
to
I
can
just
pull
in
one
of
them.
Yeah.
E
I
mean,
I
think,
if
you
want
to
do,
is
ring
battery
in
there
just
make.
You
know,
make
note
align
cervix
the
basis
for
my
position
align
uses
and
if
you
just
have
to
do
some
API
argument,
marshaling
our
normalization
right
to
deal
with
it
inconsistencies
there.
Then
we
can
do
that
and
then
either
deprecated
or
make
one
like
move,
move
eventually
to
having
the
same
API.
But
I
always
have
you
have
to
support
the
old
one
which
everyone's
getting
change
for
backwards
compatibility,
but
right.
A
C
That's
that
something
I
wanted
to
build
in
so
Ryan
I
think
had
some
code
in
his
menu
which
made
the
menu
viewport
aware.
So
that
would
like
switch.
Could
you
have
opposite
direction,
and
I
would
like
to
look
into
that
and
see
if
we
can
like
whenever
this
position
align
stuff
is
done.
I'd
like
to
look
into
that
see
if
I
could
do
that,
because
I
feel.
If
that
could
be
done,
then
this
could
work
a
lot
better
on
mobile.
C
If
we
had
the
viewport
aware
thing
because
then
you
could,
you
could
switch
automatically
and
you
I
think
you
need
to
have
some
server,
some
sort
of
an
attribute
that
forces
it
to
stay
where
you
want
to.
If
you
need
that,
but
that's
something
I'm
trying
to
build
in
so
yeah.
That's
why
these
are
things
I'm
still
working
on
another
thing
that
it
does
is
actually
delegate
delegates
as
like
an
event
delegate.
C
So
when
you
set
this
up
under
the
hood,
there's
actually
a
delegate
attribute
which
you
can
set,
or
it
sets
itself
the
document
and
then
essentially
the
selector
gets
passed
in
as
a
filter.
So
it
only
really
listens
to
one
listener.
It's
not
like.
C
It
listens
to
all
your
all,
the
well
tipsy's
on
the
page
right,
so
that
that's
kind
of
neat,
as
far
as
like
setting
events
on
it
right
now,
I
do
that
through
the
show
on
and
hide
on,
attributes
which
are
by
default
mouse
over
touch
start
and
focused
and
hide
on
his
mouth
out,
touch
end
and
blur.
But
you
can
essentially
come
to
customize
to
show
on
a
specific
event
that
occurs
on
a
specific
node.
A
C
Mm-Hmm
and
then
I
rode
this
wide
I'll
pop
over
widget,
which
extends
tipsy
and
just
mixes
in
which
its
standard
module.
So
now
you
have
accessed
a
header
content
body
content,
footer
content,
except
I
kind
of
just
made
it
so
that
you
can
do
that
through
HTML
attributes,
as
well
through
data
header
data
body
data
footer,
so
that
essentially
just
makes
you
really
easy
to
just
make
popovers
a
question
I
had
was
regarding
organization
I'm,
not
sure.
C
If
this
gallery
popover
module
should
live
in
a
separate
repo
or
if
it
should
live
with
the
wat
with
the
tips
gallery,
tipsy
repo,
because
that's
how
it
is
now
like
right
now,
I
have
gallery
tipsy,
J
us
and
then
gallery
popover
Jas
and
just
in
the
metafiles
I'm
just
saying,
or
so,
in
the
build
files.
I'm
just
saying
making
you
build,
but
I
don't
know
how
the
loader
would
pick
this
up.
I
don't
know
if
it
would,
as
a.
C
Yeah,
I
remember
you
talking
with
this.
I
think
it
could
yeah.
I
think
it
could
be
I
like
so
what's
the
functionality
that
you
need
exactly.
F
B
F
C
A
C
Yeah,
well,
the
persistence
is
pretty
easy
to
do,
because
you
can
do
that
through
the
events,
but
it
doesn't
actually
create
a
wrapper
around
a
widget.
Well,
actually
could
because
you
can
actually
pass
in
up
whatever
your
content
is.
You
can
actually
pass
in
the
node
instance.
So
if
you
pass
in
the
widgets
bounding
box,
then
technically
should
be
able
to
do
it.
Yeah.
C
Yeah
yeah
I
have.
C
Right
well,
yes,
you
could
do
that
as
long
as
said
body
content
accepts
a
note
which
I
think
it
does
yeah.
E
C
Yeah
and
I
also
wrote
unit
tests
and
stuff
furred
I
had
some
issues
getting
coverage
numbers
through
yogi
because
until
it
was
on
the
gallery
like
you,
we
couldn't
get
access
to
it
or
something
like
that.
So
I'm
still
figuring
that
stuff
out
so
I
do
have
tests.
But
this
is
on
a
separate
branch
which.
C
C
C
C
So
I'm
not
sure,
that's
something
that
I
ran
into
like
yesterday
yeah
the
other
question
I
had
was
regarding
organization
of
code
in
these
repin,
the
shifter
like
in
the
source
files.
So
right
now,
I
have
pop
over
and
tipsy.
Can
you
guys
who
see?
This
was
just
like
a
little
small
is
very
small.
Ok,
one
sec!
Let
me
switch
back
up
to
the
other
branch.
C
Okay,
so
I
better,
yes,
okay!
Well,
I
just
have
tipsy
and
pop
over
right
now
in
the
same
branch,
mostly
because
I'm
lazy,
but
this
seems
like
a
bad
practice,
because
usually
you
want
something
like
tipsy
base
and
tipsy
here.
That
makes
sense,
but
to
have
completely
different
names
here
seems
wrong.
Is
that
something
that
I'm
doing
yes?
I'm
gesh
change?
You.
E
Know
I
think
if
you
could
look
at
like
how
brian
has
the
smugmug
gallery
repo
setup,
where
they
have
multiple
modules
in
that
one
in
their
own
gallery.
Repo,
okay,.
C
E
E
C
Older
okay,
that
makes
sense.
Oh
yeah,
that
should
be
easy.
The
other
thing
I
ran
into
which
was
causing
me
grief,
was
the
whole
pulling
down
the
CSS
file
automatically,
because
it's
in
skins,
so
I
think
I
thought
I
had
this
set
up
correctly,
like
I
I
had
like
a
tipsy
Court,
which
is
right
now.
C
Blank
and
I
had
a
skins
folder,
this
probably
too
small
for
us
to
see,
but
it's
essentially
the
skins
folder
with
the
Sam
Fuld,
underneath
it
and
then
I
have
a
gallery
tipsy
skin
dot,
CSS,
but
it
just
doesn't
get
pulled
in
and
I
have
skinnable
set
to
true
doesn't
get
pulled
in
for
some
reason.
I'd,
like
put
it
in
explicitly
so
I,
feel
the
documentation
around.
This
is
a
little
like
week
for
our
stuff.
C
Yeah
yeah,
so
exactly
so,
the
stuff
in
here
should
really
be
prefixed
with
skin
Sam.
Again,
that's
just
me
being
lazy,
but
it's
not
getting
pulled
in
is
the
is
the
bigger
issue
it
does.
It
does
the
stuff
we're
like
it
finds
this
and
it
takes
whatever's
in
here
in
ads,
whatever
some
tipsy
core
to
it,
and
it
does
all
the
bundling
up
fine,
but
it
just
doesn't
pull
it
down.
That
might
be
because,
like
something
an
issue
with
loader
locally
and
Yui
3
gallery
locally,
maybe.
C
The
other,
the
other
question
I,
had
now
that
I'm
making
a
bunch
of
these
gallery
modules
is
how
to
keep
like
how
to
do
versioning
for
gallery
modules.
Cuz,
let's
say
I
work
on
some
code
over
the
week.
That's
not
getting
pushed
out
until
the
next
gallery
build,
but
I
need
to
update
my
history
md
before
that
gallery.
C
Build
goes
out
so
I've
been
keeping
I've,
been
keeping
this
version
tag
which
doesn't
really
have
any
meeting
in
my
meta
Jason
file
and
is
referring
to
that
in
my
history
MD,
but
I,
don't
know
if
there's
a
better
way
to
do
that.
A
A
C
C
That
done
my
plate
to
add
custom
animations
I'm,
trying
to
figure
out
how
to
I
feel
I
don't
wanna,
really
pull
in
like
drink
the
transition
module
to
do
that.
I
I
think
it
would
be
nice
if
that
was
just
done
through
CSS,
maybe
like
you
should
have
like
I.
Don't
know,
I
might
have
like
some
different
CSS
animations
and
you
could
like
it
could
just
toggle
the
class
name
based
on
the
animation
you
want,
but
I
don't
want
it
to
pull
in
transition
that
make
sense.
C
C
A
E
We
for
people
stuff
yeah,
so
we
we
realized
that
the
version
of
handlebars
which
you've
released
in
390,
which
was
the
tag
in
the
handlebars,
the
main
fork
of
handlebars
this
project,
it's
the
version,
that's
tagged
through
one
na
ho
ho
RC
3
and
that's
the
one
we
have
in
390
there's
a
issue
with
that.
One
related
to
nested
essentially
nested
blocks.
E
So
you
like
to
an
if
and
then,
if
you
have
like
an
estimate,
if
there's
like
a
problem
there,
so
it
essentially
makes
it
unusable
and
we
we
got
word
that
there
is
problem
there
and
it
seems
to
be
completely
fixed
in
the
most
recent
version
of
handlebars,
that's
put
been
pushed
to
mpm,
which
still
is
under
the
RC
3
name,
but
has
a
tag
for
npm
as
10,
10
and
so
locally.
When
I
updated
to
that
does
seem
to
have
fixed
the
issue.
E
So
right
now
we're
exploring
how
to
best
get
that
fixed
out
research.
They
have
two
options.
One
is
a
patch
release
for
390,
which
the
only
change
on
it
would
be.
This
updated
and
of
ours.
The
most
recent
published
version
of
antivirus,
the
other
option,
is
to
release
a
391
off
of
death
master,
which
will
include
some
other
bug
fixes
along
with
it
and
so
we're
currently
evaluating
which
of
those
two
is
going
to
be
the
best
thing
to
do.
No,
but
either
way
that
specific
fix.
We
want
to
get
out.
E
So
to
Jenny's
been
asking
about
the
diff
that
the
diff
is
pretty
small
I
have
to
check
my
email
to
see
what
exactly
it
was
I
think.
A
E
That's
supposed
to
exist
on
the
robber
object
instance,
and
if
it
doesn't,
there's
kind
of
silently
fails
now
and
or
Italy
says
it
cannot
call
a
function
on
undefined,
which
is
not
a
valuable
error.
So
we
made
that
error
addition.
Works
will
sit
by
saying
hey,
it's
basically
missing
this
function
at
this
name.
So
that's
a
very
minor
change
that
essentially
it
has
no
impact
which
should
still
be
tested
because
there's
a
lot
of
change
to
limit
and
then
we
have
wideout
tree.
E
The
data
structure
now
is
adding
a
source
property
to
its
event,
facades,
where
it's
adding
on
extra
properties
to
the
event.
So
that's
an
enhancement
and
a
very
easy
to
understand
enhancement
as
it's
just
adding
a
new
source
property
and
there's
a
lot
of
changes
there.
And
then
we
have
charts
and
graphics
which
have
gone
through
some.
E
E
A
E
Yeah,
so
that
there's
kind
of
a
there's,
some
secondary
issues
around
dealing
with
this
whole,
integrating
handlebars
into
Yui,
which
is
that
the
people
who
are
maintaining
the
handlebars
project
do
not
provide
changelogs
for
what's
changing
from
release
release
and
they
don't
have
a
consistent
story
about
tagging
versions.
So
it's
really
like
trying
to
go
into.
E
Out
where
the
cuffs
are,
you
don't
really
know
where
if
they
labeled
every?
Thank
you
very
much
easier.
Hotels
like
that
sort
of
thing,
so
there
are,
are
a
significant
amount
of
changes
from
in
handlebars
from
the
version
we've
included
with
390
compared
to
what's
the
most
recently
published,
12
m
p.m.
but
they
are
both
labeled
as
RC
3.
E
So
that
one
of
the
things
I
plan
to
do
is
add
this
particular
failure
case
that
we
found
this
regression
on
to
our
test
suite
just
very
easy
to
add
in
and
it's
a
pretty
basic
test.
So
it's
not
like
interest
ad
and
I
will
catch
if
this
regression
pops
up
again
and
protect
us
against
it
and
then
I'll
something
to
contribute
that
same
test
case
but
formatted
for
handlebars
test
suite.
But
essentially
you
can
test
the
same
logical
path
to
their
project
so
that
this
would
this
would
have
been
call.
We.
E
A
E
Yeah,
so
that
that's
another
issue,
which
is
how
we
want
to
do
the
version
number
to
be
essentially
a
cember
but
work
on
the
CDN,
and
maybe
it
may
end
up
just
being
something
we're
on
the
CDN.
It's
something
like
390,
390
patch,
one
like
similar
to
it
will
be
as
a
PR
one
we
can
patch
one,
but
then
I'm
mpm.
We
would
want
to
do
I.
Believe
me
if
I'm
wrong,
plus
patch14
ass,
ember,
4m
p.m.
for
patches
yeah
as
we
descended
if
I
do,
dash
or
minus
it
means
before
I
mean
I
know.
E
E
After
so
I
think
we
need
to
do
a
plus
yeah.
What
will
we
do
that
differently?
On
the
CDN
yeah?
We
don't
know
if
the
CDN
will
work
with
a
plus.
So
you
probably
not
want
to
risk
that
and
probably
just
deal
with
that
inconsistency
there,
because
it's
really
just
a
divider
that
we're
talking
about
so
I
feel
I
definitely
don't
feel
confident
in
the
CD
and
handling
pluses
in
the
URL
correctly,
especially
when
they're
in
query
strings
I,
don't
think
that's
less
pretty
dangerous!
E
I,
don't
think
less
is
the
right
thing
you
use
them
yeah,
so
I
think,
but
plus
it
is
a
is
actually
used
when
you
have
a
version
that
specified
XYZ
and
then
that's
the
future
release
so
say
if
we
had
a
those
minus
its
referred
to
the
past,
so
so
like
basically
check
this
out.
So
it's
like
talking
about
this
is
so
build
version
may
be
denoted
by
name
and
higher
higher
precedence
of
the
associated
no
okay
yeah.
So
it's
plus
most.
B
E
I
got
it
sorry,
yeah,
sorry
for
the
confusion
to
mix.
So
what
we
concern.
What
we
could
do
is
I
mean
I,
don't
even
know
how
we
use
the
package.
Jason
version
number
for
PRS
I
would
have
to
look
to
see
what
we
do
up.
Dave,
don't
lead
us
for
that
yeah,
but
whatever
we
do
for
that,
we
can
probably
just
replace
the
word
PR
of
catch
and
be
okay,
and
we
know
the
CDN
can
handle
that
as
a
string
and
just
be
fine.
Definitely
don't
want
to
put
the
plus
in
there.
For
this
now.
E
3914
exactly
so,
we
can't
call
it
391,
even
if
it
has
one
change
meet
us
okay,
we
can't
call
it
391,
even
if
it
has
just
the
patch
or
iPad
change,
which
is
just
changing
as
little
as
possible.
We
can't
call
favor,
we
call
it
3,
90,
plus
1
I,
know
I'm,
just
trying
to
my
front
front
Alaskan
evening.
It's
just
I've
lost
I
know.
E
The
reason
is
because
it
will
signal
the
people
that
you
can
upgrade
this
because,
for
you
know
without
worry,
because
the
only
thing
that
has
changed
is
just
we
fixed
the
specific
bug
and
handlebars
that
has
nothing
to
do,
have
a
simpler,
that's
just
based
on
our
own
condition:
okay,
yeah
and
then
the
reason.
Why
is
because,
essentially
you
can
think
of
our
version?
Numbers
is
being
one
off
from
cember
or
we
we
don't
bump
the
three.
When
we
do
API
changes
we
bump.
E
You
know
a
minor
number,
look
at
the
82
and
nine
yeah
okay,
so
we're
kind
of
like
one
off
okay
from
though
that's
what
I'm
not
gonna,
drag
right
on.
I
think
that's
that
nice
is
logical.
So
that
way,
when
we
do
a
391
people
expect
there
to
be,
you
know,
oh
I
may
have
to
pay
attention
to
the
release
notes
before
I
upgrade
this,
whereas
for
the
other
one
it's
going
to
be
like
hey,
just
as
specifically
we'll
need
to
fix
this
handlebars
issue
pop
it
in
replace
your
390
version
number
with
patch
one.
A
A
E
Will
end
up
in
the
next
release,
that's
on
the
side
of
saying,
don't
use
that,
but
so
I
think
I
placeable.
Both
ways,
though,
like
you
can
say,
oh
yeah,
these
changes
are
going
to
be
in
here
in
two
weeks:
anyways
just
throw
them
in
now
or
hey
you,
shoes
are
gonna,
be
here
in
two
meets
anyways,
let's
hold
off
on
them.
E
E
A
well
luckily,
all
these
two
points
that
we're
talking
about
so
the
charging
graphics.
We
know
we
just
need
to
test
Safari
6
run
through
ellen
test
to
be
totally
confident
with
them,
because
the
other
things
that
have
changed
charge
does
not
depend
on.
Similarly,
there
is
nothing
else
all
these
up
all
the
changes.
The
other
models
are
all
leaf
modules,
so
there's
no,
they
have
no
dependents,
so
nothing
else
has
to
be
tested
because
nothing
else
depends
on
those.
So
we're
lucky
in
that
regard,
that
our
testing
surface
here
is.
E
G
If
these
are
all
eat,
then,
like
take
the
lazy
wanna
list
check
in
mm-hmm,
you
know
what
testing
was
done
ahead
of
that
check
in.
It
was
a
great
that
we
shouldn't
really
even
H.
We
tested
right
yeah.
E
E
That's
browser
dependent
or
at
least
cold.
I
was
being
browser
dependent
and
assuming
it's
been
lifted
for
traveling
combos
and
that
sort
of
thing
that
would
cause
ie6
d'ivoire.
Then
there
is
very
little
risk
or
that's
right.
That's
for
the
robber
14
for
the
lazy
modulus.
It
is
also
just
logical
changes
like
basic
java
script
changes
no
domina
that
so
none
of
these
things
are
don't
think.
There's
any
dominant
lining
yeah,
so
they're
very
well,
they're
all
low
risk
to
begin
with
and
they're
leaf
nodes,
and
we
can
easily
nanuet
test
them
all
three
chests.
E
Like
have
the
you
know,
the
various
becker,
really
things
or
the
engines
are
different,
and
that
seems
worth
thank
ya,
especially
since
you
said
like
we
could.
We
could
use
ready
to
pump
through
all
those
tests
locally
on,
like
my
machine
writings,
partly
to
safari
and
do
that
in
five
minutes.
Yeah.
E
G
E
We
definitely
can
I
mean
we
could
we
could
take
an
approach
to
say
like
hey.
These
are
all
you
know.
We
know
these
are
all
like
basic
java
script
changes.
We
can,
you
know,
be
smart
about
how
we're
testing
them
and
test
that,
like
an
IE
6,
which
will
hit
any
trailing
common
issues
and
given
IE
runtime,
do
it
in
Firefox
dude
at
a
WebKit
from
time,
do
it
in
oj
s
and
then
give
it
a
thumbs
up
right,
especially
the
tree
ones.
E
A
E
A
E
E
I
E
G
J
A
J
E
G
E
No,
no
sorry
I
should
go
more
clear.
The
version
of
handlebars
published
as
the
handlebars
nvm
module
is
them
updated
to
the
newest
one
they
don't
yet,
but
they
still
call
that
thing
internally,
rc3
like
if
you,
if
you
download
that
code
and
then
you
say,
require
a
handle
and
you
say
in
a
bar
stop
version
it
will
pronounce
in
three.
Even
though
it's
a
completely
different
code,
then
then
what
rc3
was
originally
released
us
so.
E
They
didn't
tell
him
it's
not
a
formal
release.
It's
an
interim
release
thing
moving
to
not
put
any
change
along
about,
so
that's
what
I
was
hard
to
figure
out
that
there
was
even
a
problem
here
and
also
why,
like
we
haven't,
run
into
it
in
a
lot
of
cases
because
we're
just
using
handlebars
directed
from
NPM
and
not
wideout
handlebars
on
our
server
side
stuff.
So
we
were
getting
the
fixed
version
and
never
been
currying.
This.
A
A
B
E
E
I
Mean
I
think
if
we
don't
want
to
do
a
patch
unless
there's
like,
maybe
just
a
resources,
yours
you
know
which
is
too
much
work.
You
know
it's
not
worth
it
yeah.
Then
we
probably
need,
and
we
don't
have
to
do
it
right
now.
We
probably
need
to
revisit
our
policy
and
process
because,
if
we're
having
to
follow
it,
yeah
language
change,
it
yeah.
B
E
E
H
A
E
A
B
G
I
H
I
E
Like
how
are
we
I
message
this
and
it
it's
moneyer
that
way
where
this
is
like?
Hey
we
don't.
They
did
big
big
thing
for
this
release.
We
updated
handlebars
and
oh,
look.
You
get
these
extra
busy
fixes
in
some
documentation
for
promises
worth
release
worry
I'm,
actually,
that's
enough
stuff
to
be
released,
worthy
and
good
to
go
and
still
were
able
to
highlight
that
the
major
thing
of
this
is:
we
fix
them
handlers
issue
by
pulling
in
the
latest
version
I'm
with
you.
A
E
G
A
messaging
thing
that
says
whatever
whatever
so
remember,
version
wants
to
a
390,
390
patch
comes
out.
That
means
everyone's
on
390
should
just
drains
that
automatically
update
your.
E
G
G
E
F
G
E
G
B
A
E
E
G
E
A
G
E
E
We
should
definitely
do
that
on
the
blog
post,
but
I
think
the
core
thing
we
saying
that
blog
post
is
like
install
NPM,
install
yui
at
stable,
like
we
want
to
have
that,
but
we
could
say
that
the
version
number
44
m
p.m.
is
a
cember
which
is
using
a
plus
and
for
URL
CDM
reasons.
The
plus
does
not
exist.