►
Description
A
We
forgot
to
start
the
recording
and
the
streaming
before
our
stand
up,
but
since
that's
the
tallest
part
of
the
meeting
and
the
part,
that's
probably
the
best
document,
it
isn't
minutes,
that's
probably:
okay.
We
have
a
ton
of
things
on
the
agenda
today,
so
we're
going
to
see
how
many
of
them
we
can
do
without
needing
to
talk
much
here
and
first
up
back
and
get
others
to
make
fast
decisions.
But
that's
not
gonna
be
possible
with
everything.
So
now
I
stop
talking
and
we
move
on
to
the
agenda,
which
now
involves
me
talking.
A
The
role
of
sense
of
CC
incentive
or
major
changes
needs
clarification,
I'm
going
to
open
that
issue
and
I
think
that
that
that
can
keep
happening
on
github
and
we
can
move
on.
I
think
the
conversation
has
gone
on.
There
is
fine,
there's
some
nuanced
differences
of
opinion
or
viewpoint
about
what
the
role
ctc
should
be,
but
we
should
tease
those
out
get
those
more
explicit
and
then
either
just
get
resolve
it
with
a
pull
request
or
else
bring
it
back
to
this
meeting
next
week.
So,
let's
move
on
revert
the
fs
changes.
D
C
The
one
that
made
the
call
back,
checking
mandatory
on
all
async
functions
in
an
FS
apparently
broke
the
planet
they're
quite
a
few
modules
out
there,
in
particular,
quite
a
few
used
by
NPM
that
call
FS
async
FS
functions
without
passing
in
a
call
back.
So
when
we
started
throwing
errors
it,
you
know
it
means
things
like
NPM
set
fail
to
work
crackers
now
this
is
it
technically
it.
It
can
be.
Viewed.
B
C
Bug
fixes
making
this,
but
as
with
anything
that
adds
error,
handling
or
angry,
you
know
the
deal
to
their
Hamlin.
We've
always
treated
it
as
being
a
fairly
sensitive
issue.
These
did
land
is
some
of
our
majors,
but
right
now
and
p.m.
one
master
is
broken
because
of
this,
so
this
PR
just
proposes
that
we
refer
to
changes
for
now,
there's
another
PR
that
is
opened.
I
believe
this
morning
that,
instead
of
throwing
an
error,
emits
a
warning
if
the
callback
is
not
passed,
so
we
have
a
couple
of
ways
we
can
do.
C
A
A
C
C
A
C
D
C
D
D
E
Known
on
the
pull
request,
78
46-
we
did,
there
are
two
changes
that
are
mentioned
and
I
find
it
a
bit
confusing
to
follow
like
what
we're
talking
about
precisely
because
the
two
changes
are
different,
and
so,
if
we
could
split
that,
maybe
into
to
a
different
pull
requests
discussing,
reverting
each
surprise
me.
Maybe
that
would
help
have
a
discussion
that
is
a
bit
less
confusing
a
very.
E
F
D
C
That
so
the
second
question
after
we
revert
so
the
reason
we
made
this
change
in
the
first
place,
or
at
least
one
of
these
was
because
it
is
a
bug
to
allow
these
async
functions,
be
called
without
a
call
back.
So
the
second
question
and
I
can
remember
the
exact
issue
or
put
new
porpoise
number,
but
the
question
is:
do
we
want
to
emit
a
warning.
D
A
D
D
I
just
have
a
quick,
although
the
graceful
at
best
morning
with
nvidia,
but
that
really
was
when
I
was
just
all
feels
compiling
either
than
the
other
day
and
the
latest
version
of
assam
when
you're
running
it
spits
out
like
10
or
15
of
those
mornings.
Those
things
like
that,
let
us
know
it's
still
morning
and
so
I
think
we
have
a
riskier
of
of
warning
to
tea
people,
just
not
just
getting
you
to
them.
D
It's
also
a
lot
of
people
of
interpret
warning
that
there
is
and
I
unable
to
tell
the
difference
just
because
there's
too
much
faculty
and
also
the
fact
that
it's
really
noisy
like,
should
we
instead
of
going
for
a
warning,
perhaps
we
should
get
a
new
version
of
you
guys
flip
it
out.
Let
that
sit
for
a
while
before
we
proceed
warning.
C
C
Well,
what
one
of
the
approaches
that
I
asked
miles
to
just
kind
of
experiment
with
and
see
if
it
would
be
viable
if
we
can't
get
the
fix
integral
SS
directly
as
we
can
slow
to
catch
on
NPM
so
that
it
catches
the
version
of
grayskull
offensive
npm
uses
within
archery.
Ok,
you
can
install
them
hinthrop,
you
yep,
lino,
ok,
so
it
I'm
not
saying
permanent
measure.
I'm
saying
oh
my
job,
this
is
operational.
C
B
C
A
A
B
A
A
A
A
D
D
A
C
G
The
windows
side
also,
some
updates
discovered
that
I
can
for
more
issues
that
were
a
few
trackers,
are
also
dependent
on
those
changes
and
it
spent
spent
from
the
new
FS
field
path.
Normalizing
the
drive
letter
to
uppercase
always
which
called
some
really
that's
the
side
effects
on
windows
because
on
our
path
resolve
and
requires
all
produce,
different
keys,
and
so
it
becomes
fairly
easy
to
get
modules
to
blow
twice
on
the
way
on
the
windows.
Side,
I
think
we're
stuck
in
a
corner
and
we
need
to
reverse
it
and
the
kid.
C
I,
don't
believe
I
gave
any
contacts
for
what
I
was
good
thing
that
the
benchmark
specifically
targeted
using
required
without
module
cash.
So
basically
the
idea
would
be
neat
and
brings
acting
old
implementation
of
FS.
Bill
path
went
bust
risk,
but
do
not
include
the
cash
option
so
that
we
would
be
able
to
change
the
implementation
under
the
hood
without
a
major
version,
change
and
I,
measured
that
through
using
require,
since
that
is
the
Vegas
use
case.
So
the
numbers
I
gave
her
cortex
in
regard
to
using
and
using
possible.
A
A
G
D
C
So,
whether
whether
okay,
I
would
like
to
see
not,
I
would
like
to
see
that
we
don't
require
the
cash
option,
but
whether
the
performance
hindrance
is
great
enough
to
allow
us
to
leave
it
out.
I
am
going
to
leave
others
to
decide
that,
like
I
can
weigh
in,
but
I
don't
have
a
strong
opinion
either
way.
G
C
F
D
C
G
B
D
C
A
C
D
A
People
want
matto
1799,
please
do
so
and
then,
if
it
is
not
landed
by
what's
today,
Davis
unleaded
by
Friday
hot
landed,
but
please
please
somebody
James
and
of
somebody
Trevor
somebody
landed
before
that.
That
would
be
my
double
you
might
hope.
I,
don't
think
what
it
feels
different
like.
We
need
to
be
more
cautious
or
something,
but
I
think
this
is.
This
is
one
of
those
things
that
required
that
we
want.
D
C
As
a
particular
point
on
this,
for
450
does
include
some
miners
because
minor
updates
to
the
LCS
branch
because
of
those
we
want
and
because
of
the
risk
of
introducing
regressions,
he
wants
to
be
very
conservative
about
making
sure
we
have
a
good
release
candidate
test
cycle
for
this,
because
he
doesn't
want
to
add
to
the
noise
around
you
know.
Oh,
we
get
a
regression
suddenly
things
that
are
totally
unstable,
Bubble
Ball.
C
C
I
have
been
kicking
this
around
we're.
Not
quite
it's
not
quite
ready
to
put
anything
down
on
paper
yet,
but
probably
within
maybe
the
next
week
I'm
going
to
try
to
get
some
ideas
posted
to
the
LCS
working
group,
and
then
we
can
kind
of
discuss
it
from
there.
So
maybe
in
a
week
we
can
kind
of
continue.
The
conversation.
A
C
C
D
C
C
B
C
B
C
A
The
fans
of
you
streaming
for
us,
okay,
all
right
alive
sister
army
else,
online,
have
any
comments
on
the
puny
code
issue.
D
C
A
D
A
C
C
Instead,
which
is
exposed
now
through
a
process
binding
the
in
the
future,
with
the
soft
deprecation
that
wouldn't
change,
the
icmp
module
would
still
be
there.
This
particular
PR
that
open
the
discussion
would
introduce
heart
deprecation,
which
moves
the
pinna
code,
module
into
an
internal
and
then
uses
a
new
sim
that
just
printed
the
deprecation
message
right.
So
when
you
require
communicated
with
Prince
deprecation
message
that
wasn't
change,
maybe
eventually
down
the
road.
If
we
decided
to
okay,
we
want
to
go
ahead
and
remove
it.
I
Other
kind
of
question
I
had
with
that
is
you
know
like
way
on
down
the
road
like
we
know,
go
past
whatever
hard
deprecation,
it's
not
available
externally,
at
all
anymore,
severely
internal
thing
for
when
people
build
without
I
see
you
you
know
is
this.
You
know
you
had
mentioned
that
there
was
a
newer
version
on
NPM,
that's
faster
and
all
this
stuff.
Are
we
going
to
continue
to
like,
like
update
the
internal
one
for
the
non
I?
See
you
people
going
forward?
Are
we
just
going
to
leave
it
sit
there?
I
would.
C
Yeah
Nikita
just
pointed
out
in
the
github
thread
that
the
Unicode
module
on
NPM
has
a
5.7
million
downloads
per
month,
which
indicates
that
people
are
already
off
of
there
using
the
new
Virginia,
the
newer
version
so
yeah,
it
I
think
it's
a
bit.
You
know
doing
the
top
deprecation.
Now,
I
think,
is
fine.
We
just
kind
of
leave
it
as
it
is.
F
F
D
C
F
C
A
C
A
G
A
Okay,
great
then
again,
unless
there's
something
else
on
that
we're
going
to
move
on
to
http,
don't
inherit
for
mocking
prototype.
Is
this
you
Brian
I.
D
Noticed
that
Chris
Dickinson
isn't
here,
but
he
was
having
it
some
concern
about
the
fact
that
we
merged
the
null
prototype
thing
on
query
string.
He
identify
the
fact
that
we
did
that,
and
it
was
surprised
by
it
and
calling
it
out
of
the
concern
so
just
to
voice
that,
on
behalf
of
him
on
Twitter,
that
was
obviously
a
deterrent
I.
C
C
Something
like
a
it's
unlikely
that
someone
have
a
to
string:
query
parameter,
name
that
it
makes
much
more
sense
to
have
a
parsing
model
that
allows
that
allows
it
instead
of
having
these
arbitrary
names
from
from
the
JavaScript
implementation
not
be
allowed.
So
I,
hey
I,
can
understand
the
concern
over
the
breakage,
but
the
query:
Street
thing
has
been
out
there
for
months:
a
change
Brian
flavor
wrong.
I
think
that
land
is
like
january
or
something
right.
It's
been
other.
It's
been
landed
for
quite
a
while,
and
we've
had
no
major
regressions
reported
from
I.
C
B
C
And
it's
like
the
only
one
we
I've
heard
it
up
and
we
with
regard
to
this
particular
PR
with
the
the
HUP.
This
is
with
regards
to
the
headers.
We
specifically
held
off
landing
this
because
we
wanted
to
get
the
input
from
the
user
community
and
so
far
all
of
the
input
has
been
well.
It's
still
like
to
string.
We'd
still
like
has
done
property,
but
everything
else
is
fine
and
let's
just
kind
of
figure
out.
C
If
we
can
do
in
those
two
right,
like
maybe
borrow
those
methods
from
the
object
prototype,
but
the
other
question
I.
The
other
thing
that
we're
looking
at
is
recent
benchmarks
for
the
map,
so
that
using
map
is
actually
faster,
surprisingly
than
using
a
null
prototype.
So
the
question
is
well.
We
can
actually
my
vision
that
step
you
say
faster.
What
do
you
mean
by
faster
like
construction
and
both
youth?
You
sentence
like
access
yet
activated,
maybe.
A
G
C
E
A
So
I
I
wonder
if
the
oh
yeah,
I
guess
I
get
my
feeling
about
this
and
I
want
those.
This
is
everyone
else's
general
feeling
is
that
when
I
went
for
rods
to
be
back,
hey
Rox
back,
it
is
hot
in
here,
but
that's
okay,
because
we're
going
to
be
out
of
here
soon.
All
these
laptops
were
all
running
anyway.
So
on
so
might
my
preference
is
that
you
know
we
do
this.
We
landed
us
get
get
get
up.
A
You
know
how
you
be
an
object
with
either
an
old
prototype
or
prototype
that
as
far
as
the
one
or
two
things
we
actually
think
people
want
/
need,
but
with
it
but
with,
but
now
that
be
a
interim
measure
with
the
goal
of
moving
to
map
implementation
assuming
performance
is,
is
similar
or
better.
How.
I
F
C
D
A
A
A
C
B
E
C
Nhp
is
extremely
unlikely,
but
just
the
fact
that
it's
a
valid
header
that
we
cannot
support
it
will
throw
an
error.
Okay,
so
I
just
sent
it.
We
collaborated
is
always
only
it's
only
an
issue
if
somebody's
cents
at
double
underscore
proto
header,
and
then
somebody
tries
to
access
something
like
to
get
a
lot
to
string
on
the
object.
Yes,
not
after
we're
putting
into
this
thing
that
is
so
previously.
A
I
Probably
not
too
much
more
than
what's
or
even
said,
I
mean
I
guess
my
preference
would
be
I'd,
be
leaning
more
towards
landing
at
just
at
least
because
it
would,
if
nothing
else,
it
would
maintain
consistency
with
how
we're
already
handling
the
query
string.
Parsing
I
mean
I,
don't
know.
B
B
D
A
All
right,
so
maybe
Brian
can
you
add,
mentioned
chris
dickinson
on
the
issue
to
try
to
like
get
his
feedback
and
and
otherwise
I
mean
you
know,
take
it
from
there.
I
guess
Julia.
A
D
A
Right
so,
let's
leave
that
on
the
agenda
in
case
it
continues
to
be
stuck,
but
I
think
we
have
a
password.
A
D
A
A
C
Spd,
a
chromium,
I'm
reference,
that's
like
what
we
should
use
going
forward,
so
we
could
write
a
PR
now
and
my
great
said:
if
we
wanted
to
we,
we
have
the
preferred
copyright
block
as
well.
It's
just
a
matter
of
like
do.
We
want
to
do
this
all
at
once.
There's
there's
also
the
matter
of
we
need
to
add
the
licensing
informations
back
into
a
bunch
of
files
where
it
was
removed.
The
joy
and
copyright
line
was
moved
removed.
A
So
Michael
and
rod
will
take
the
tail
figure
out
who
will
do
what,
when
it
happens
and
all
that
sort
of
stuff
does
anybody
the
remoter
in
the
room,
have
any
questions
or
anything
to
say
about
the
legal
advice
on
license
and
copyright
blocks
in
code?
Or
can
we
safely
move
on?
Can
we
actually
remove
this
one
from
the
agenda
for
know.
D
E
C
A
E
So
Mike
I
think,
if
I
understand
correctly,
Michael
Dawson
put
it
on
the
agenda.
Gallon
wagon
and
I
suggested
that
maybe
it
would
be
something
that
we
should
discussing
it
with
in
the
tsc
necessary
within
that
the
CDC,
but
maybe
I'm
missing.
Something
I
knew
better
also
was
in
that
discussion,
so
maybe
feller
has
something
to
add.
H
Yes,
it's
just
shorten
all
about
it.
Basically,
there
are
some
postmodern
related
projects.
That's
probably
my
wish
to
be
under
either
noches
github
organization
or
some
other
organization,
and
that's
basically
the
question:
can
they
be
either
I?
Think
I'm,
not
Michaels.
Also
I,
don't
know
if
Franco
has
any
sense,
so
I
think
Michael
is
not.
A
H
A
D
A
C
That
is
functional,
we're
passing
all
of
the
parsing
tests
and
all
the
two
of
the
center
tests
that
are
provided
by
the
what
WG
URL
the
two
senators
were
not
passing
or
intentional.
They
are
minor
and
non
consequential
failures
that
I
think
actually,
some
of
the
browsers
don't
pass
as
well.
The
poor
quest
can
constantia
review,
there's
quite
a
bit
of
C
C++
codes
that
need
just
second
set
of
eyes
on
to
to
kind
of
go
through
and
do
the
review
what
it
is
functional.
The
one.
C
Performance
of
the
parsing
algorithm
isn't
quite
as
good
as
the
oral
parse
implementation
is
currently
that's
because
the
URL
parts
implementation
take
some
shortcuts,
which
means
it
fails.
160
of
the
test
of
the
parsing
tests
you'll
make
that
up
with
the
Uni
to
it.
Okay,
so
there
is,
you
know
so,
there's
a
performance.
C
C
J
C
C
Parsing
get
the
cottage
satisfied.
Ok,
that's
what's
good,
it's
slower,
I
know:
I
am
opposed
to
a
global,
but
I
mean
it's
more
correct.
Honestly,
in
this
case,
I,
don't
even
care.
If
it's
slow
I
mean.
D
C
A
registered
and
/
via
the
provision
scheme
list
scheme
list
and
I
am
amazed
at
how
many
ridiculous
schemes
and
it
was
able
to
parse.
It
was
kind
of
stupid,
but
yeah
I,
know
I,
think
it
handles
everything
better
than
our
current
implementation
and
like
I'm,
fine
with
that
I
just
want
to
global
yeah.
D
D
A
So
I'm
going
to
interject
myself
here
to
say
that
this
is
probably
a
long
conversation
and
we're
already
over
the
one
out
we're
already
five
minutes
over
and
we're
probably
not
going
to
come
to
a
decision
today,
you're
not
looking
for
that
right,
you're,
just
looking
to
raise
awareness,
get
more
eyes
on
the
EPS,
and
sooner
or
later
we
have
to
vote
on
the
EPS
and
say
yes,
this
is
what
I'm
going
to
do
or
no.
This
is
now
we're
going
to
do
and
do
it
some
other
way.
C
C
A
Afford
it,
maybe
we
can
do
is
only
it's
going
to
just
try
to
put
approximate
times
for
things
that
that
makes
making
the
agenda
more
work,
but
if
there's
only
a
few
items,
I
might
be
able
to
take
a
stab
at
that
or
something
okay,
so
yeah,
okay,
so
civilized
will
probably
have
a
similar
conversation
like
this
one
next
week,
just
sort
of
like
brief
updates,
hello
back
and
forth
and
then
bigger
conversation
in
two
weeks
is
the
hook.
Yep.
F
C
A
With
that,
is
there
anything
anybody
wanted
to
add
to
the
agenda
at
the
beginning
of
the
meeting
that
didn't
get
added
before
we
move
on
to
Q&A
forum,
public
fora,
okay,.
E
A
A
D
A
Some
of
questions
right
you're,
like
watching
one
watcher-
oh
alright,
cool
well,
in
that
case
we're
just
going
to
move
on.
I
think
I
might
want
to
do
something
about
this
Q&A,
we're
obviously
being
much
more
functional
than
we
ought
to
be
if
they
were
answering
revised.
Questions
before
Q&A
arises:
okay,
because
we've
gone
over,
I'm
not
going
to
list
off
the
upcoming
meetings
on
the
recording
by
the
time
the
recordings
posted.