►
Description
A
B
The
real
path,
or
at
this
point,
lack
thereof
with
the
discussion
looking
like
it's
heading
towards
implement,
revert
and
implementing
the
fictional
TV
I'm
holding
off
on
my
PR.
There
are
two
permits
in
there,
though
I
still
think
should
land,
regardless
just
some
benchmarks
and
I'll.
Yes,
next
to
his
truck
Richard.
C
All
right
so
I've
been
tweaking
the
CTC
governance
docs
that
hopefully
Bolin's
following
along
with
probably
want
to
talk
about
that
in
the
section
at
the
end,
I
on-boarded
Andhra,
who
did
a
bunch
of
timer
performance
work,
he's
a
he's
a
collaborator
now,
so
actually
he
might
not
be
a.
He
might
not
be
a
bad
person
for
you
to
pay
Julian,
although
I'm
not
sure
that
general
timers
is
his
jam,
but
I
might
be
worth
finding
out
working.
C
I'm
gonna
do
prince
jay
wesley
next
and
after
that,
I'm
out
of
nominees,
so
I'm
open
for
all
their
ideas.
If
you
know
of
anybody
doing
great
work,
who
really
ought
to
be
on
board
it,
let
me
know
and
other
than
that,
just
doing
the
neverending,
Sisyphean
whack
Amole
of
illuminating
flaky
tests,
I
believe
Jeremiah's
here
now,
so
we
can
go
back
to
him.
If
there's
anything
to
report.
E
Okay,
it's
just
rolling
rolling
to
me
so
previous.
If
there's
anything
arising
out
of
this,
that
you'd
like
to
comment
on
Pluto
I've,
been
wanting
to
move
through,
we
talked
about
dropping
support
for
visual
strategic
reserve,
they've
been
still
near
Denver
again,
this
week's
we've
talked
about
again
the
real
path
Akane
over
again
object
prototype,
inheriting
on
that
from
for
the
HTTP
headers,
that's
a
Legionnaire
again,
and
he
450
is
only
a
generic
again,
and
so
is
what
do
we
do?
Url
you
have
James
this
week.
He.
C
E
F
F
B
Well,
it
looks
like
that
looks
like
everybody's,
leading
towards
just
doing
a
full
revert
early
said
mostly
parcel
river
until
we
can,
until
the
dvds
implementation
is
fixed.
I
still
find
that
unfortunate
on
the
Linux
for
basically
non
window
size,
because
we
have
a
fix
for
everything
but
the
window
side.
B
But
if
that's
everybody
later
what
he
wants
it,
though
I
made
comment
that
I
believe
we
should
look
into
the
suppose
if
performance
issues
and
why
we
need
cash
and
see
if
see
how
performance
does
without
cash
and
if
it
shows
to
be
just
fine
and
I
said
we
leave
cash
out.
Leah's
also
points
out
that,
if
we
can
leave
that
out
that
we
can
change
the
implementation
underneath
without
any,
without
needing
to
do
a
Mickey
revision.
G
Yeah,
basically
I'm
pretty
much
a
Greenville
Trevor
that
I
find
it
a
bit
unfortunate
that
we
would
have
to
do
a
full
revert
for
the
unique
systems
like
sagol
made.
A
good
point
about,
like
only
we
working
for
windows,
would
double
the
API
servers
and
the
surface
for
bags,
which
might
be
something
we
don't
want
to
do.
G
Also
yeah,
like
I,
don't
know
if
we
have
any
benchmarking
numbers
for
like
requiring
projects
with
a
large
number
of
files
like
how
much
is
the
star
of
time
there
how
performances
performances
that,
because
I
believe
that
is
the
primary
function
of
the
cash
to
work?
Well,
there
am
I,
don't
know.
If
note
club
has
any
benchmarks
miles
mentioned
that
they
are
using
the
Cashin.
It
seems
like
something
where
we're
benchmarking
could
make
sense.
H
We
do
have
two
benchmarks
that
run
every
night,
but
they
and
the
one
is
basically
I,
think
a
require
cached
and
a
require
non
cached,
but
I
don't
know
how.
Well
they
mapped
to
the
scenario
you
said,
like
you
know,
a
project
using
a
large
number
of
files
or
not,
though
we
may
like
at
one
point,
we
saw
a
fairly
large
in
Crete.
You
know
improvement
in
one
of
those
benchmarks
and
I'm
wondering
if
maybe
we'll
see
the
corresponding
decrease.
If
we
do
this
reversion.
H
I
E
We
go
that
route,
then
we
need
to
make
sure
that
Trevor's
work
on
resolving
the
e
loop
stuff
is
integrated,
because
then,
because
that's
one
of
the
overlapping
areas
where,
if
we
left
it
as
it
is,
then
we
would
have
iluu
being
resolved.
Okay
on
windows
been
done
on
non-windows,
so
we'd
have
to
still
have
all
that
stuff
pulled
in.
G
A
B
B
E
The
my
concern
here
is
LTS
we're
heading
an
internal
use,
version
6
and
it
we
can't
leave
this
in
this
situation.
For
that.
So
that's
why
I'm
in
favor
of
a
you
know
close
to
full
reversion
I'm,
actually
don't
mind
whether
we
go
with
that
with
or
without
the
case
option.
It
seems
to
me,
including
it
would
be
the
logical
path,
because
then
we
get
back
to
original
status
and
then
we
can
work
on
an
improved
version
for
version
71
words
just
simply
so
that
we
have
stability
going
into
LTS.
B
That's
how
this
originally
came
up,
if
you
all
the
way
back
to
that
first
issue,
I
had
just
a
side
comment:
oh
well,
why
don't
we
just
use
the
like
in
node?
Why
don't
you
just
go
ahead
and
use
you
know
access
and
sell
Paul
things
like
well.
Let's
just
do
that
in
lib
EV
and
then
forget
fair.
Who
was
somebody
finally
implemented
in
Liberty
and
now
we're
here.
A
Saying
so
much
problems
now
of
it?
Is
it
out
of
the
question
to
revert
the
the
interface
but
also
create
another
public
interface
that
goes
to
the
lib
UV
method?
Because
I
am
you
know
if
we're
talking
about
stuff
between
the
two
LT
esas
we're
about
to
upgrade
to
1.9?
So
it
wouldn't
be
completely
crazy
that
we
could
expose
that
as
a
minor
update
to
give.
You
know
like
broccoli,
for
example,
access
to
that
interface
to
do
faster
resolution,
but
not
make
that
the
default
resolution
for
anything
else.
We're
also
slow.
E
J
E
Connect
before
they
could
just
they
could
try
and
do
something
on
their
end.
I.
Imagine
that
maybe
this
has
come
up
because
they
wanted
to
try
and
fix
performance
on
area.
But
if
we
push
back
and
said,
look
we
quit
just
for
this
version.
We
can't
then
they
might.
You
will
need
to
revisit
on
there
I.
D
We
were
doing
on
at
this
point.
I
mean
I,
probably
have
been
a
favor
of
reverting
like
two
weeks
ago,
because
this
is
been
going
on
for
a
really
long
time,
and
we
should
like
worry
about
fixing
this
stuff
before
that
other
stuff,
just
acknowledge
that
it's
still
like
a
problem
in
the
state.
It's
in
envy
for
two
for
users
that.
I
Let
me
agree:
we
should
not
be
compromising
correctness
or
performance.
It's
been
broken
for
a
while
now
in
terms
of
the
bringing
back
the
cash
argument.
I
would
be
more
if
they
were
keeping
the
removal
of
the
cash
argument
so
that
this
is
not
a
breaking
change,
so
it's
pretty
original
fix
was
an
intended
breaking
change
in
the
removal,
cash
argument
and
an
untended
making
change
in
the
windows
issues
and
the
loop.
B
I
B
Yeah
well,
whatever
thing
in
like
benchmarks,
I
mean
we'll
also
have
to.
This
is
really
the
whole
to
benchmark
because
of
SSDs
now
more
popular
than
was
originally
implemented,
be
cashing.
It
felt
like
Colonel
little
cashing
I
fine,
like
oh
I'll,
start
writing
my
desk
and
first
it's
like
holy
crap
I'm
doing
500
days.
B
A
second
I
realized,
oh
no
I'm,
not
hard
drive
or
the
Colonel's
just
cashing
it
up
or
like
holy
crap,
suddenly
I'm
able
to
access
my
files
so
much
faster
phone:
oh
I'm,
not
the
curls,
kashtam,
all
right
so
properly,
Baines
parking!
This
is
going
to
be
a
bit
of
a
pain
we'll
have
to
like
create
a
huge
number
of
things,
require
them,
dump
them
and
create
a
whole
bunch
more
things
and
require
them.
If
we
want
absolute
fresh
like
the
times
without
cached
right,
if.
F
E
Is
why
I
am
in
favor
of
a
fuller
version?
Just
simply
because
if
we
start
you
know
debating
the
cash
option,
then
we're
going
to
postpone
this
for
another
month
or
two.
We
need
action
to
happen
now,
it's
a
it's
a
breaking
change,
but
it's
like
it's.
It's
instill
an
object,
so
it's
actually
not
going
to
break
anyone's
code.
It's
just
going
to
break
expectations.
If
they're
using
a
new
option,
which
you
know
they
can
only
using
version
six
anyway,
so
it's
not
going
to
actually
break
people's
code
in
you
know.
D
G
E
A
A
B
E
A
E
Ok
back
to
github
for
now
and
let's
try
and
see
if
we
can
magically
get
more
resolution
before
next
week
and
sound
like
we
have
anything
to
do
a
vote
on
for
this
right
now,
but
I'm
github.
Okay!
Next
is
the
v50
proposal
miles
that
you'll
join,
say
anything
unless
or
shall
we
move
on
from
that
I?
Don't.
A
Think
that
there's
much
to
be
said
but
I
it
doesn't
seem
like
there's
been
tons
of
people
putting
eyes
on
it.
So
you
know
I
had
emailed
the
folks
over
at
NPM.
I
have
not
heard
a
response
back
yet
so
maybe
offline
rod
you
and
I
can
just
talk
about
how
we
can
improve
this.
What's
going
on
right
now,
yep.
E
A
E
L
C
E
M
E
C
J
F
M
Right
so
I,
so
my
understanding
of
the
summary
of
the
issue
was
that
Bieber
we
wanted
to
go
with
a
null
prototype
and
the
the
general
sense
I
saw
got
was
that
eventually
we
want
to
move
to
a
map
down
the
road,
and
I
want
to
make
sure
that
we
don't
have
to
go
through
22
changes
that
are
big
and
if
there's
something
we
can
do
so,
if
you,
if
the
goal,
is
to
go
to
a
map.
What
is
preventing
us
from
doing
that
now,
because.
D
E
M
Through
getters
and
setters,
we
might
be
able
to
get
there
and
the
reason
actually
I
want
to
explore.
This
is
because
the
the
headers
is
a
source
for
a
lot
of
D
ops,
and
not
idea
is
a
lot
of
mega.
Morphic,
ICS
and
and
map
is
purpose
built
for
it.
So
I
think,
there's
performance
implications
of
this
that
are
visible,
not
necessarily
in
in
micro
benchmarks,
in
in
real
world
world
workload.
This
might
have
more
interesting
behaviors
that
I
want
to
explore.
Okay,.
E
That's
fine,
let's
not
make
it
move
on
this
other
this
week,
but
I
just
want
to
recap.
The
last
week
we
did
leave
off
on
the
suggestion
that
will
retreat
this,
just
like
with
query
string,
which
is
removing
all
like
just
a
null
backed
object
and
removing
all
things.
So
that
was
the
proposal
that
we
left
off
on.
So
if
you
wish
anyone
wants
to
have
a
look
at
this
during
the
week,
we
can
come
back
to
it
next
week,
yeah.
C
I
think
I
just
inject
one
thing:
real,
quick,
I
think
that
I,
you
know
we
were
talking
about.
You
know
putting
on
an
object
and
removing
the
you
know
the
underbar
underbar
proto,
and
things
like
that
and-
and
the
map
of
course,
is
the
ideal
but
I
mean
brian.
Is
it
correct
to
assume
that
the
the
object
change
you
know
putting
in
an
object
with
nope,
you
make
it
keeping
it
as
an
object,
but
removing
the
you
know.
C
The
prototype
chain
is
a
relatively
straightforward
thing
and
the
maps
will
not
be
and
there's
no
reason
that
that
the
object
can't
be
a
step
towards
the
maps.
If
that
will,
if
that
will
work
or
if
there's
performance
benefits
there,
and
then
you
already
have
all
the
tests
in
place
to
like
check.
You
know
the
things
that
that
that
the
simple
fix
will
fix.
D
This
is
probably
a
better
idea.
There
is
an
issue-
it's
probably
like
four
years
old
on
the
old
issue
tracker.
It's
like
a
lot
of
people
like
asked
for
a
proper
API
for
this
for
and
I
tried
to
look
at
that
like
early
last
year,
I
think
there's
a
bunch
of
problems
and
that
resolved
a
rat
revolve
around
how
we
deal
with
implicit
headers
with
actually
making
api's
for
this.
D
But
if
we're
going
to
like
look
at
changing
how
all
this
fundamentally
works,
we
should
really
just
make
an
API
for
it,
because
that
way,
we
won't
really
have
this
problem
of
having
to
adjust
as
legacy
thing
that
you
know
hopefully
can
eventually
go
away.
Maybe
we
are
stuck
with
some
form
of
it
with
up
for
the
meantime,
but
that
should
really
be
the
ideal
path
forward
here,
maybe
not
just
ideal
I
think
it
probably
should
be
seriously
investigated.
E
E
E
So
the
upcoming
meetings
that
we
are
aware
of
next
week
is
next
Thursday
ctc
meeting
next
from
sorry,
everyone
next
Wednesday
Oct,
see
you
next
Thursday
is
TFC
and
although
its
node
summit,
so
few
of
us,
will
be
out
of
action,
so
keep
it
up.
Updates
on
that
build
working
group
is
on
in
three
weeks
time.
Lts
is
on
next
week.
Apparently,
Diagnostics
is
on
the
third
of
April
adversary,
August,
and
we
don't
know
first
of
all
or
API,
but.
H
A
H
E
Okay
also
miles
I'd
like
to
talk
about
moving
and
shifting
LTS,
two
or
three
weeks,
three
weeklies
level,
but
we
can
talk
about
that
in
there
ltsp
by
and
so
that's
it
for
public
meeting
yeah.
We
have
some
private
business
to
move
on
to,
but
thanks
everyone
for
listening
and
you
can
listen,
please
we're
going
afterwards.
Thank
you
for
taking
an
interest
right
now.