►
From YouTube: Node.js Build WG meeting - Feb 19 2019
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
Yes,
I'll,
just
repeat
what
I
said
in
there
in
just
in
private
there,
that
Mac
stadium
has
renewed
their
sponsorship
of
the
build
working
group.
So
they're
obviously
happy
enough
with
the
arrangement
that
we
have
with
them.
We
do
a
little
bit
of
promotion
for
them
just
to
let
people
know
we're
using
them
and
so
they've
renewed
their
sponsorship
for
another
12
months.
So
the
28th
of
February
is
when
it
comes
up
again
for
review.
A
C
B
C
C
A
Right
I
mean
I,
think
I
think
the
focus
on
this
launch
was
on
you
know
the
smaller
prototype
I
think
that
I
think
there
is
interest
in
using
it
to
prove
out
the
work
they're
doing
on
the
new
website
as
well,
and
if
that
ends
up
being
positive,
I
think
then
you
know
at
some
point
they'll
say:
hey:
we've
we've
taken
it
to
a
certain
extent.
We
think
this
looks
good
for
the
new
website.
How
do
we
you
know?
What's
what's
the
transition
path
from
the
website
we
have
now
to
to
that?
A
But
I
guess
you
know,
my
perception
is:
is
it's
still
early
early
sort
of
days
right,
like
nobody
is
committed
to
that
particular
infrastructure?
It's
more
of
a
let's
prove
it.
You
know,
use
it
as
a
place
to
prove
out
the
concepts
and
move
the
new
website
forward
and
then,
when
it
gets
to
a
certain
stage
there
you
know
we
would
need
to
have
a
more
formal
relationship
with
the
build
working
group
to
figure
out
how
it's
going
to
you
know
be
maintained,
going
forward.
B
A
I
mean
I
think
it's
it's
kind
of
one
of
those
things
where
we
don't
have
a
ton
of
time
and
and
so
I
don't
think
it's
like.
We
can
jump
in
and
be
super
active
until
we're
sort
of
until
they
vetted
it
and
once
they
vetted
it
then
yeah
we
better
make
sure
there's
some
good
a
good
amount
of
time
for
a
handover
and
transfer
and
whatever
and
I
guess
Raphael
are
you?
Are
you
sort
of
keeping
in
the
loop?
Are
you
in
the
loop
at
all
or
just
sort
of
Wow.
C
A
A
B
A
B
That
that
just
turned
into
a
let's
alibis
the
things
that
we
do,
and
so
there
was
this
idea
that
rich
and
I
would
spend
some
time.
Looking
at
that
great
do
so
do
a
little
sprint
of
really
cataloging
it
so
that
we
see
what
we're
doing
what
resources
with
we're
using,
not
not
just
the
sponsorship
stuff
but
people,
time
right.
D
B
Yeah
I
mean
I
have
to
be
honest,
I
just
haven't
spent
much
time
on
know
in
the
past
couple
of
months,
I've
just
been
too
distracted.
Things
right,
my
band
I
did
do
I,
don't
know
if
we
want
move
on,
but
I
just
spend
a
bunch
of
time
on
the
raspberry
pi's
there's
an
issue
in
there
that
Raphael
made,
and
maybe
we
could
discuss
yeah.
A
D
B
A
A
B
Come
both
of
these
both
of
these
are
their
long-standing
two
dues
for
me,
they're,
really
a
I
already
done
pretty
important,
because
a
lot
of
it
is
about
stuff
that
you
know.
These
are
things
that
are
the
kinds
of
things
that
mostly
in
people's
heads,
mainly
my
head,
and
we
need
to
get
them
out
anyway.
Mm-Hm.
A
B
A
A
Or
we
could
set
another,
you
know
we
could
set
it
I
can't
remember
if
we
do
it
every
two
or
three
weeks
now,
I
think
it
might
be
three
weeks.
B
A
A
B
B
So
you
know,
I
can
I
can
win
the
winter
when
they
drop
off,
I
fix
them
up
and
all
that
sort
of
stuff.
It
does
seem
to
be
at
least
a
couple
that
I'm
suspicious
of
because
they
they
you
know,
there's
certain
what
like
I
recognize
them
by
number
and
name
and
there's
a
couple,
but
just
keep
on
coming
back
up
and
then
even
when
I
replace
the
SD
cards,
and
so
that
makes
me
suspicious
I'm,
very
integrity
and
so
we've
got.
B
We've
got
fourteen
of
them,
two
of
them
reserved
for
bills,
they're
in
release
that
that
will
eventually
change
when
we
stop
doing
because
we're
doing
cross
compiled
now
for
renewal
releases.
So
when
we
phase
out
one
of
them,
I
know
if
it's
good,
we
do
that
in
eight.
Maybe
we're
doing
that
in
a
already.
So
maybe
we
don't
even
need
them
six
and
we're
at
least
doing
the
sixth.
B
B
So
if
we,
if
we
cranked
it
up
to
say,
can
paralyze
across
ten,
it's
still
the
initial
cost
of
the
checkout
and
and
set
up
of
a
file
system
that
that's
that's
quite
a
cost
for
these
machines,
so
it
would
speed
it
up
a
little
bit,
but
it's
it's
really
diminishing
returns
for
greater
paralyzation
because
of
that
cost,
but
the
type
the
testing
suite
is
getting
bigger
and
so
we're
faced
with
you.
Do
we
want
to
think
about
retiring
him?
Do
we
want
to
think
about
doing
something
more
creative
with
the
test
suite?
B
Maybe
paring
it
down
for
this
engine,
because
this
is
not
just
a
problem
for
us
replay
that
this
is
going
to
be
more
and
more
of
a
problem,
because
where
are
we
at
the
tests?
We've
increases
and
if
you
keep
growing
quite
fast
because
of
the
courage
meant
for
coverage
and
every
new
feature
needs
multiple
tests.
B
D
C
B
We
just
I
did
put
a
chart
in
my
last
comment
on
the
issue,
which
is
number
one:
six
seven
seven
in
build
and
it's
got
percentages
of
the
different
architectures
and
v6
is
at
0.08
percent
of
total
downloads
in
comparison
to
arm
v7,
which
is
point
two.
B
B
B
I
commented
there
I
just
don't
know
how
even
how
they
interpret
that
because
you
can
like
it.
That's
that's,
that's
more!
That's
still
more
than
arms
64,
which
I'm
still
surprised
that
I'm
64
just
hasn't
been
as
successful
as
I'm
hoping
for
kind
of
a
stuff,
and
even
you
know,
I'm
64
is
even
greater
than
the
IBM
platforms,
so
yeah.
C
B
A
A
D
It's
for,
and
probably
they
probably
can't
provide
much
health,
but
the
hardware
team
working
group,
whatever
it
is,
has
recently
been
resuscitated.
As
Michael
knows,
yeah
I
don't
know
if
I
don't
know.
If
there's
any
insight
that
those
folks
might
have
us
to
you
know
if
these
sorts
of
things
are
being
used
much
or
not
with
node,
but
someone
on
there
might
have
some
weird
insight
that
we
don't.
B
Ok,
absolute
numbers,
so
v7
is
let's:
let's
pick
the
18th
of
February
mp7
is
2300
downloads
on
v6
is
1000
download.
A
B
C
B
A
A
B
C
A
B
Another
thing
another
strike
against
dropping
support
for
on
v6
is
that
it's
really
hard
to
compile
it
yourself
from
b6
cross-compiling
is
annoying
like
really
annoying
getting
it
right
for
your
because
there's
so
many
variants
of
arm
b6
you've
got
to
get
just
the
right
instruction
set
and
supported.
You
know
flags
great.
A
B
C
B
It
is,
but
I
have
not
been
a
pretty
live,
got
pretty
broad
support,
the
way
they
compiled
and
I've
not
heard
of
them
not
working
on
another
round
v6
that
somebody's
tried,
I
had
not
seen
a
report
for
that.
That's
a
similar
cake
question
son,
an
OS
ones
like
people
constantly
complain
hey.
This
is
not
generics
on
iOS
stuff.
This
is
this.
A
A
C
B
Was
raised,
but
we
have
in
the
past
bumped
into
things
with
three
six
witnesses
and
I
think
that
was
a
long
time
ago.
I
guess.
C
C
C
A
C
I
talk
so
I,
there
is
an
honest
and
a
bit
of
a
tangent.
There
is
a
community
driven
effort
which
potentially
would
probably
be
me,
help
I
archive
to
create
a
community
build
of
note
for
Linux
x86,
like
a
a
binary
that
doesn't
have
that
is
not
blessed
there.
They
they
have
this
project,
they
want
to
have
it
embedded
and
they
have
a
bunch
of
hardware
that
they
need
it
for.
C
A
C
C
C
B
B
B
A
B
B
C
There
is
a
set
of
tests
which
I've
collected
from
the
debug
builds,
which
are
just
slow,
and
their
impact
on
the
total
coverage
is
less
than
1%.
So
we
round
them
only
once
a
day
and
it's
it's
a
nice
I
think
it
was
in
in
total
time
it
was
like
something
like
15%
of
total
test.
Time
is
in
that
list,
which
for
now
is
just
skipped
on
the
debug.
C
Look
at
this,
something
you
know
stochastic
like
say:
if
the,
if
your
load
on
the
raspberry
pi
it's
a
certain
threshold,
then
we
start
dropping
tests
and
then,
when
the
loan
subsides
Iran,
we
could
check
master
and
see,
like
anything
wrought
for
for
now,
with
suits
that
we
moved
into
this
strategy.
It
hasn't
been
a
horrible
and
rich
might
I,
don't
know,
also
has
an
opinion
about
that
removed
all
the
internet
tests
or
the
Palmer
debts
and
throw
the
benchmark
tests.
We
move
to
testing
only
once
a
day
and
slow
debug.
D
C
B
C
D
C
B
So,
let's
just
there's
more
costs
here,
so
this
is
not
just
about
learning
peers,
the
slower,
the
more
tests
we
have
to
run
and
the
slower
things
are
gone
longer
it
each
release
takes
so
when
we
have
these
batches
of
releases
like
security
releases,
and
we
need
to
queue
up
all
these
tests.
If
they
are
each
taking
an
hour
just
to
run
basic
tests,
then
you've
got
to
add
sitting
on
top
of
it.
B
Release
has
become
more
and
more
costly
and
difficult
and
frustrating,
and
you
know
prone
to
it
prone
to
infrastructure
problems
that
exacerbate
things
so
the
further
we
go
down
this
without
having
a
proper
solution
or
maybe
biting
the
board
and
saying
we're
dropping
support
the
more
web.
We
are
locking
in
those
costs.
So
if
we,
if
we,
for
instance,
just
leave
our
own
v6
for
a
node
version,
12
in
three
years
of
test
runs
that
are
at
least
40
minutes.
B
We're
locking
those
costs
in
for
three
years
so
we're
going
to
be
really
really
careful
about
that,
and
maybe
maybe
we
do
need
to
rationalize
and
say
it's
time
to
to
cut
bait
and
run
I.
Wonder
if
I
wonder
if
we
could
clear.
Maybe
Twitter
is
the
best
place
to
do
this,
getting
their
jeaious
to
put
out
a
tweet
asking
for
feedback,
if,
like
we
did
with
it
for
x86
I,
think
we
did
some
I
think
if
we
drop
down
v6
in
brackets,
Raspberry
Pi,
one
yep.
B
C
C
C
B
A
B
Tried
for
so
long
to
get
that,
though,
so
here's
another
thing,
the
problem,
a
problem
for
them
is
that
raspbian
is
currently
built
for
arm
v6.
So,
even
though
it
runs
on
v7
and
64,
it's
built
for
them
lowest
common
denominator.
So
that's
still
supporting
in
raspbian
I
bet
they'd
love
to
get
off
that
because
they
cross
compiling
things
all
the
way
down.
There's
pre-assemble
I
haven't
heard
any
rumblings
of
them.
I've.
B
C
D
B
B
C
B
D
B
D
A
D
B
C
D
A
C
D
D
A
A
A
A
C
A
C
C
C
C
C
So
I'm
sure
all
the
release
channels
would
be
happy
very
happy
to
do
that.
Major.
It
hasn't
come
to
discussion
yet
because
we
were
I.
Don't
we
incompatible
right
here
once
we
are,
we
could
say
will
be
Python
3
pattern.
2
will
be
experiment
because
at
the
end
of
the
year
it's
gonna
go
anyway,
and
we
have
end-of-life
escape
hatch
right.