►
From YouTube: Node.js Foundation Build WG meeting 2016-04-26
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
Yes,
foundation
bill
working
group
meeting
of
haverhill
26
present.
We
have
myself
Alexis,
compiler
and
squishy
and
flatten
you
on
Bergstrom,
Michael,
Dawson
and
miles
boring.
So
as
usual,
we
will
start
by
getting
to
a
quick
stand
up.
I
will
try
to
take
notes
in
it
anywhere
else.
I
want
to
help
help
clay.
A
So
the
order
of
the
document-
okay,
I'm
first
so
I
happen.
I've
done
a
bit
of
work
related
to
build,
try
to
run
Canadian.
The
gold
mine
on
Windows
was
objective,
was
to
see
how
Windows
fares
and
also
to
figure
out
where
it
will
take
to
get
support
in
CI
for
Windows
runs
so
the
first
part.
The
compatibility
is
a
working
progress
and
it
doesn't
look
too
bad,
but
what
I
figured
out
is
that,
can
I
nibble
my
only
provided
success
versus
fail
for
the
whole
module?
A
It
doesn't
actually
give
you
a
breakdown
of
what
tests
fail
or
whatever
visa,
so
that,
then
I
assume
we
need
to
add
support
for
flaky
modules,
which
canonical
mind
already
has
in
order
to
specify
okay.
This
module
is
known
to
fail
on
windows,
so
we
don't
block
the
least
because
of
it
and
but
miles
was
telling
me
that
the
support
is
not
its
global
is
not
/
platform.
So
that's
something
a
little
bit
heading
yeah.
B
A
A
D
Okay,
not
too
much
to
be
honest,
but
a
few
things
at
least
I
fractured
the
the
freebsd
scripts
for
the
lynchers
an
hour
wats
to
to
improve
the
resiliency
for
restarting
and
monitoring.
I
saw
that
there
was
a
slight
issue,
but
you
are
done.
I'll
be
working
on
like
improving
the
parallel
job
running
situation,
but
from
the
bill
read
well
the
Nokia
312.
D
So
we
can
run
tests
and
linting
in
parallel
I'm
looking
at
making
that
applied
to
make
as
well,
but
it's
most
often
irrelevant
seeing
how
we
use
C
cash
work
for
most
of
I
removed
the
poverty
stuff
for
even
having
in
our
ass
for
skits
for
a
good
while
now,
which
essentially
pulled
up
github
keys
from
get
up
and
put
it
into
a
dress
host
authorized
feast
on
slave
the
Russian
Ali.
Here
being
that
we
moved
to
the
shared
test,
release
in
bra
benchmark
and
so
on,
keys
that
are
available
through
the
shared
secrets.
D
So
the
way
to
add
users
now
is
by
delegating
them
with
the
gbg
keys
to
the
secrets.
Repo
I
did
some
minor
refactoring
of
the
engine.
X
config
there's
been
a
few
pull
requests
I've
merged
with
regards
to
improving
some
some
other
stuff
and
then
genetics,
and
while
doing
so
I
did
some
cell
refactoring
I
have
a
gland
of
that
yet.
But
it's
do
everything
there's
also
a
few
other
engine,
X
config
changes
from
the
node.js
org
team,
I
I
need
to
merge,
but
I'm.
Looking
at
that
and
I've
been
reviewing
it.
D
D
Guess
we'll
talk
about
it
later,
but
I
just
want
to
know
what
options
we
have
I've
been
working
somewhat
with
the
github
bought
team,
which
is
a
relatively
new
group
of
people
working
to
create
a,
but
does
you
know
all
kinds
of
work
on
PRS
and
issues
in
the
know,
Gia's
repo,
but
hopefully
also
has
the
ambition
to
grow
more
functionality,
integration
with
Jenkins
and
so
on.
So
I'm
very
interested
in
that
and
I'm
very
glad
to
see,
see
it
hub
and
seeing
how
the
bill
group
has.
D
D
Now
that
run
out
of
RAM,
which
is
pretty
weird
because
I've
never
seen
it
prior,
but
what
I
did
was
just
essentially
it's
lat
born
four
gigs
up
swamp
on
each
leg,
so
I
need
to
put
that
intangible
and
as
late
as
this
morning,
we
also
room
also
run
into
issues
with
another
build
slave
early,
spilled
slave.
That
was
a
result
of
Jenkins
I'm,
not
sure
what
to
do
about
that.
D
It
doesn't
happen
very
often,
but
it's
just
one
thing:
we
need
to
think
about
long-term,
hopefully
well
by
hopefully
I
mean
that
long-term
someone
else
will
solve
it
for
us,
but
we
just
need
to
have
an
eye
open
for
slaves
that
they
their
flag
down
but
they're
online,
and
you
need
to
force
a
disconnect
and
wait
the
grace
time
for
it
to
work.
It's
it's
really
tricky,
but
yea,
though
that
was
the
reason
that
release
job
was
holding
up
this
morning.
B
I'll
try
and
talk
closer
to
the
laptop,
so
I,
also
adding
moving
the
pbc
big-endian
machines
over
to
a
mu
2
from
fedora,
so
that
we
can
use
them
in
the
release
of
this
and
the
other
related
worked.
But
he
ain't
testing
in
the
node
tree
so
have
the
initial
jobs
up
and
running
once
tonight
and
investigating
a
couple
of
issues
that
we've
seen
out
of
those
first
ones
being
on
our
where
and
all
the
worse
and
test.
B
When
you
were
going
to
tell
so
I've
excluded
those
it
finally
did
spend
a
little
bit
of
time
creating
some
new.
Let
me
test
jobs
along
the
way
and
then
next
is.
B
C
I'm
good
so
yeah,
like
really
stuff
I,
was
doing
some
discussions
last
night
about
how
to
speed
up
canary
in
the
gold
mine
by
doing
some
aggressive
caching
and
not
doing
as
much
grabbing
online
resources
I'm
going
to
be
looking
into
seeing.
If
we
could
do
some
of
that
this
week
and
then
I
also
want
to
get
a
few
of
those
features
in
that
will
help
support
various
platforms,
specifically
making
flaky
modules
specific
to
platforms
and
also
being
able
to
set
up
modules
to
be
skipped
on
particular
platforms.
C
A
C
B
A
But
if
you
work
within
all
sources,
you
could,
theoretically,
with
the
versioning
in
there.
C
E
A
B
Of
the
earlier
meetings,
I've
taken
the
action
to
create
that
straw,
man
and
I
just
added
it
in
because
I
don't
think
we've
got
any
comments
or
discussion
the
issue
itself.
So
more
of
a
reminder-
and
maybe
if
we
have
time
we
spend
some
time
talking
about
what's
error.
E
A
B
D
D
A
D
B
It's
as
permanent
as
any
of
us,
I've
access
terms
of
our
grounds
has
been
permitted
great
yeah
yeah,
because
that's
the
sort
of
the
difference
between
its
ongoing
versus
you
know.
Temporary.
It's
like
well
we're
gonna
they're
going
to
need
to
do
this
one
you
know
going
forward
and
therefore
will
decide
to
trust
on
this,
as
we
would
trust
the
screen
I.
So.
A
A
D
D
Is
there
anything
you
know,
can
I
get
access
to
try
things
out
and
I,
just
don't
see
like
permanent
access
for
testing
purposes
being
relevant
and
then,
if
we
look
at
how
the
bill
group
is
moving
and
growing
or
shrinking,
it's
pretty
constant,
so
so
yeah
I
guess,
there's
no
conclusion.
Just
more
inside
yeah.
B
I
guess
the
other,
the
other
alternative
would
be
to
they
were
willing
to
give
ongoing
access,
but
in
a
know
they
can
log
in
they
have
their
own
user
account,
but
not
necessarily
through
the
Cherokee
in
that
would
be.
That
would
be
more
restricted,
I
think
that
wouldn't
cover
all
the
case
that
could
be
another
class
of
an
ongoing
access
but
restricted
versus
ongoing
access
to
luda
access
on
the
machine
effectively.
What
the
guy
here,
you
should
give
you.
A
D
A
B
B
A
C
A
In
the
Jenkins
startup
script
is,
this
is
a
secret
that.
D
Not
you
mean
to
identify
the
slave
yeah
I,
don't
think,
that's
an
issue
per
se
seeing
how
each
slave
neatza
it
not
necessarily
night,
be
secret
match.
But
you
know
it's
limited
the
deal.
What
I
peace
can
access
the
Jenkins
master
is
limited
so
like,
even
if
you
have
the
secret
key,
you
wouldn't
necessarily
be
able
to.
You
know,
emulate
a
slave.
Nobody.
A
D
D
D
You
know
five,
ten
minutes
if
even
but
but
like
that
use
case,
if
at
least,
if
I
look
at
the
track
record,
I
think
I'm
the
one
probably
giving
out
post
axis,
which
it
doesn't
happen
very
often
you'd
it
like
once
a
month
and
its
own.
Most
often
you
know
collaborators
that's
been
around
for
a
year
and
they
need
access
to
a
specific
machine
and
I
have
a
like
a
text.
D
I
tell
them
like
don't
install
packages,
or
if
you
need
to
do
anything,
you
tell
me
or
stuff
like
that,
I
temporarily
out
there
their
pub
key
to
the
slave
and
then
tell
me
when
it's
done
and
then
I
revoked
it
and
also
reconnected
to
Jenkins.
If
you
don't
have,
if
we
don't
have
your
done
and
see
for
that
type
of
slave
I,
just
say
like
you
know,
have
a
look
at
the
Jenkins
and
don't
try
to
you
know,
make
it
suffer
too
much.
But
it's
just
like
try
to
be
nice
and
build
on
trust.
B
Right
so
I
think
that's
a
long
way
or
less
along
lines.
What
I've
done
in
the
past
to
they
mostly
pin
that
you
know
somebody
that
I
know
we
last
day
is
the
ax
machine
investigating
issue
that
I
think
we're
saying
like
what's
written
down
here
at
básicas
collaborators,
you
just
gettin
brother
access.
E
B
Things
that
are
listed
there
in
terms
of
you
know
locate
their
collaborative
/
decide
if
they're,
not
a
collaborator,
you
know,
have
they
been
involved
in
working
group,
pledge
that
we
know
them
or
the
you
know
the
nodejs
community.
We
know
they
are
and
press
them
or,
as
I
think
you
know,
other
people
mentioned
be
poorest
know
that
they
work
for
a
company
in
stock
because
there's
consequences
to
limit
ETA.
You
know
where
there's
others,
operators
who
will
vouch
for
the
midnight
release.
B
A
B
D
Yeah
I
guess
a
question
here
is
I
mean
if
this
is
ongoing
and
happening,
we
should
make
a
list.
First
of
all,
we
should
have
an
audit
book.
We
should
just
add
a
log
somewhere
check
it
into
the
build
repo,
where
we
just
say:
user,
xscape
user
y-axis
to
this
machine
between
x
and
y
or
something
like
that
and
all
and
I
guess.
Also.
The
second
question
is
like:
who
do
we
need
to
keep
check
of
who
are
allowed
to
delegate
access
I.
B
A
B
E
B
A
B
B
B
A
D
Sure
I'll
try
to
keep
it
short,
so
the
problem
is
stud.
Well,
it's
growing
a
note,
but
the
gist
is
we
don't
have
enough
testing
on
OS
X
and
the
current
platform
we're
testing
on
which
is
unbox
er,
which
previously
also
gave
us
freebsd
testing
they've,
pretty
much
told
us
that
you
should
look
at
moving
elsewhere.
D
Tired,
look
tat
options
in
terms
of
hosting,
obviously
the
best
state
would
be
someone
you
know
company
stepping
up
and
saying
hey
we
care
about
the
releases
on
OS
X
and
we're
happy
to
chip
in,
but
I've
also
looked
at.
You
know
commercial
services
online
in
case
we
need
to
spend
some
of
the
money
we
have
so
yeah
I
guess.
The
problem
is
that
you
just
don't
have
enough
people
that
care
enough
about
this
to
push
it
forward,
but
I
think
the
kind
state
is
documented,
pretty
well.
D
D
A
D
That's
all
in
all
so
I
mean
the
problem
is
twofold:
we
don't
have
enough
hardware
and
then
the
tricky
part
is
that
once
you
have
the
hardware
you
need
to
find
a
software
apple
won't
provide
you
with
any
old
any
other
versions
than
current
or
prior
so
10
6,
7,
8
9.
If
we're
interested-
and
we
have
to
find
ourselves,
you
know
dig
up
an
old
USB
or
something
like
that,
and
that's
the
good
part
about
at
least
mack
stadium.
It
gives
us
the
option
of
choosing
older
operating
systems.
A
D
D
A
D
D
It's
a
bit
unfortunate
I
mean
I,
really
want
to
see
a
company
stepping
here
that
uses
know
it
on
OSX
a
lot
I
think
your
developers
in
general
does
but
it'd
be
great
to
get.
You
know,
company
support,
and
hopefully
you
know,
improve
the
state
of
of
the
bill
group
by
having
more
vistas
being
interested
in
the
development
of.
A
D
Now
I
can,
I
can
try
and
write
something
a
bit
less
specific
now
I
mean.
Obviously
my
finest
findings
in
terms
of
options
might
be
a
bit.
You
know
limited
to
mack
stadium,
whereas
I
think
something
more
generic
would
probably
be
beneficial
if
someone
wants
to
provide
hardware
or
similar.
So
so,
if
you're
suggesting
that
I'd
create
an
issue
in
the
notary
boy,
I'm
happy
to
do
so.
Yeah.
A
D
Like
that's
I
mean
we
just
always
our
sorry.
No
j6
will
drop
seven
and
eight,
but
the
problem
is
that
we
still
haven't
defined
what
we
support
for
you
know:
0
10,
12
45,
it's
still
undefined,
so
you
know
before
I
think
from
from
a
buildable
perspective
before
we
come
forth
with
our
recommendation.
This
is
what
we
should
support.
Officially
we
at
least
need
to
test
which
we
we
don't
at
the
moment.
So
it's
a
bit
of
you
know
chicken
egg
situation,
but
I
like
it.
D
D
Yeah
we
could
do
that.
I
mean,
I
know
the
big
one
was
fired,
10
526
when
they
switch
to
you,
know
the
lov
until
Shane
and
moving
forward.
It's
obviously
just
improving,
so
I'm
happy
to
you
know,
revise
the
list.
If
that
helps
absolutely
it's
just
that
we
as
a
bill
group
I,
think
we
need
to
can
I
say
this
is
what
we
recommend
and
if
it's
10-7
or
e
10-8,
that's
all
good
for
me,
I
I'm,
not
on
no
expert
in
OS
X
either
do.
B
C
So
we
just
bumped
the
minimum
OS
X
version
in
v6.
That
was
like
one
of
the
one
of
the
polls
that
came
in
at
the
end.
Let
me
just
verify
exactly
what
number
it
is
so
we've
bumped
the
mac,
OS
x
development
target
two
to
10
7
as
a
minimum
target.
I
believe
that
v8
5
has
dropped
support
for
seven
and
eight
because
they
no
longer
receiving
security
updates
from
Apple
at
all,
I
believe.
E
D
A
C
B
I
hold
it
like,
you
know,
from
our
IBM
bills,
we
believe
in
sporting
Teddy
Bob,
so
it
doesn't
really
help
with
the
lower
level
ones.
I
guess
the
key
key
thing
for
the
LTS
is
like
people
actually
use
it
for
production
as
OS
X
in
production,
and
that's
where
you
know
LTS
work
is
intended
if
its
people
using
it
on
their
laptops,
it's
pretty
hard
to
not
be
on
a
later
novel,
is
what
I
understand.
C
D
Yeah,
and
also
now
that
six
is
just
out
I,
think
you
know,
checking
the
NBN
repo
checking
our
repo
and
unchecking
homebrew
for
issues
with
guards
to
you
know,
deployment
targets
might
be
relevant
if
there
is
like
a
lot
of
people
saying
hey
I'm
on
tonight
or
whatever
they're
on.
You
know.
That
might
be
a
good
indication
for
us
to
consider
in
terms
of
LTS
yeah.
C
One
thing
also
that
I
think
is:
we
can
do
this
incrementally
I
think
getting
10
10,
10,
11
and
10
9.
Automated
testing
going
should
be
the
top
priority
and
then
we
can
start
handling
the
other
ones
like
if
it
becomes
necessary.
There
was
like
an
issue
that
came
up
in
the
v6
release
just
before
it
came
out
saying
that
apparently
a
test
was
failing
on
10
11,
you
know
I,
usually
don't
update
OS
X
until
you
know
that
thing's
been
outbid,
like
usually
I'm,
always
staying
one
release
behind.
D
Yeah,
I
mean,
I
think,
is
finding
finding
that
company
that
actually
uses
OSX.
You
know
deeply
cares
about
it
and
different
versions
of
no
node
and
OS
X
would
be
a
good
fit.
I
mean
then
I'm
happy
to
do
hardware
because
they
would
likely
have
all
the
resources
available
to
us.
Otherwise
it
can
get
tricky
yeah
because
we're
not
going
to
find
the
software
right.
B
E
A
Alright,
we
could
try
to
reach
out
to
Michael
and
see
if
some
of
the,
if
we
could
put
an
article
out
on
the
blog
as
a
call
forum
out
on
the
hosting
issue,
an
intelligible
version.
Support
I've,
been
where
we
suggest,
starting
with
lieutenant
11,
sounds
good
bye
week,
also
broaden
the
issue
to
release
the
collaborators
and
see
if
to
have
any
further
opinion.
Even
there
are
more
sex
users
with
an
opinion.
B
B
B
D
B
D
B
B
D
Yeah
I,
just
wanna
I,
think
like
if
I'm
the
only
guy
pushing
this.
I
don't
think
visibility
will
be
that
great.
So
if
you
know
if
there
was
a
working
group
that
knows
how
time
to
buy
this,
or
if
there
are
channels
we
have
up,
may
might
make
it
more
visible,
that'd
be
preferable.
I'm.
C
B
C
The
one
thing
I
do
think
is
worth
bringing
up
here,
though.
I
know
that
rod
was
pretty
adamant
about
wanting
to
not
take
foundation
money
for
this
if
possible,
and
to
try
to
have
it
something
like
be
something
coming
from
the
community.
So
I'd
want
to
make
sure
that
you
know
like
rods
not
around
right
now
and
I.
Don't
want
this
to
be
something
that
seems
like
it's
going
around.
You
know
his
wishes
on.
D
A
E
A
A
Great
so
1053
here
the
next
issue
is
accept,
accept
terms
of
use
for
linux.
One
machines,
issue,
number
386,
yeah,.
B
I
really
is
that
this
is
a
reminder,
so
jus
already
sent
me
his
email
saying
yes,
he
accepts
and
it's
just
I
need
go
in
and
Roger
to
do
the
same.
E
B
Okay,
I'll
go
back
to
my
email
and,
if
I
don't
find
it
all
sending
email,
just
saying
hey,
can
you
send
it
again
otherwise
yep?
I
find
it
sorry,
sorry
for
not
fer
for
losing
it.
No.
B
A
A
They
said
that
the
city
of
port
project
was
having
trouble
testing
that
we
ease,
especially
on
windows,
so
we
built
this
Jenkins
job
and
there's
a
machine
donated
by
Microsoft
I
physical
machine
in
the
city,
port
device
and
so
there's
a
job
to
test
the
master
branch
of
serial
port
every
day
and
it's
running
now
in
the
know.
Jsc
are
and
I
just
wanted
to
get
a
feedback
on
whether
that's
the
right
place
to
do
it.
If
it's
okay
to
be
part
of
little
Jessie
I
or
if
we
should
create
a
separate
master
for
it.
A
A
A
E
D
B
A
B
A
B
A
D
Mean
I
think
there's
this
a
point
to
be
made
about
the
distinction
between
the
bill
group
and
we
have
a
good
call,
Jenkees
admins
at
the
moment,
which
you
know
if
you're
in
that
group,
you
have
admin
access
to
both
of
our
Jenkins
environments,
meaning
you
can
do
all
changes,
you
can
mess
with
slaves
and
it's
kind
of
implied
that
you
have
access
to
most
machines
as
well.
So
I
can
see
how
it's
you
know
there
might
be
a
distinction
to
be
made
between
them,
but
yeah
I,
don't
really
see
a
problem.