►
From YouTube: Node.js Foundation Build WG Meeting 2016-07-19
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
And
we're
on
no
jst
foundation
build
working
group
meeting
for
the
theme
ninth
of
july,
two
thousand.
Sixteen.
A
Their
gender
somewhere,
let's
give
nobody
way
somebody
got
eaten
dating
from
them.
It's.
B
C
B
Who
okay
octa
startin
so
compared
to
our
since
last
time?
I
wasn't
present
last
time
so
I'm
going
to
a
bit
over
the
last
four
or
five
weeks
at
least,
but
I'm
I'm
mostly
been
doing
worked
with
the
the
arcing
or
the
unencrypted
machine,
as
we
call
it,
which
is
intended
to
host
our
an
unencrypted
downloads.
B
We
have
like
older
ssl
clients,
aren't
able
to
download
nodejs,
and
our
sting
has
also
been
a
feature
requests,
a
long
studying
page
request
from
from
bigger
organizations
to
send
to
have
a
simpler
mechanism
of
downloading
our
stuff,
so
we're
at
the
final
end
of
the
run
in
terms
of
testing
it
out.
We've
gotten
some
feedback
and
we
have
a
machine.
That's
currently
been
manually
kept
up
to
date,
but
we
have
a
script.
B
That's
meant
to
be
deployed
shortly,
so
the
only
change
since
the
last
time
we
spoke
about
it
was
that
I
have
had
a
shot
with
a
few
people.
That's
giving
us
feedback
and
the
main
feedback
is
that
the
end
of
life,
which
we
suggested
being
first
of
January
2020,
will
be
extended
till
the
first
of
January
2022.
B
That
means
that
we're
adding
it
to
EXO
years
to
it
most
because
our
sink
is
one
of
those
things
that
you
said
and
you
deploy,
and
you
forget,
and
the
feedback
was
in
general
that
it
was
too
short
for
them
to
to
commit
larger
mirrors
to
that.
So
if,
unless
anyone
else
has
any
objections
to
extend
it
with
two
years,
I
think
that's
what
we
should
run
with
I've
also
done
some
PVC
work
with
Michael.
We've
had
a
fair
amount
of
problems
again
with
keeping
the
machines
alive.
B
It
seems
like
it's
a
combination
of
the
vm
host
and
the
software
that
is
running
in
the
stack.
That
seems
to
be
the
issue
this
time.
The
good
news
is
I.
Guess
Michael
should
talk
more
about
this
because
he's
done
most
of
a
job
but
yeah
that's
ongoing
work
with
this
and
I've
done.
Some
minor
stuff
I've
also
been
working
on
replacing
the
top
test
suite
with
j
unit.
B
Similarly,
to
what
miles
been
doing
in
canary
in
the
gold
mine,
I
I
still
don't
have
a
pull
request
up,
but
it's
almost
there
and
this
will
probably
be
the
biggest
relief
for
our
CI
and
hopefully
we'll
make
the
CI
run
fast
again.
I
I
think
this
is
one
of
the
biggest
issues
we
have
with
memory
consumption
from
the
Jenkins
host
point
of
view.
B
So
the
problem,
the
main
problem
at
least
throw
a
lot
of
what
I've
done
identified
with
our
tap
plugin
for
Jenkins,
is
that
it
stores
all
output
and
build.xml
and
on
every
time
you
load
a
build
XML,
which
is
pretty
much
every
page.
Imagine
going
to
test
commit
Linux.
It
opens
up
the
blob
parses,
the
blog
and
converts
it
into
J
unit,
which
is
the
only
input
that
the
Jenkees
test,
like
admin
can
display.
B
C
B
A
B
B
Will
likely
I
mean
it,
what
we
do
is
we
write
it
to
file
and
then
Jenkin
comes
and
slurps
it
up.
Hopefully
it
will
be
stored
separately,
but
even
if
it
isn't
it's
just
a
smaller
xml-
and
it's
I
mean
hopefully
miles
knows
more
about
how
it's
actually
stored,
but
from
my
testing
it's
just
instant.
You
like
it's
no
parce
time
to
just
open
a
test.
So
it's
a
huge
improvement
in
terms
of
how
quick
it
takes
to
load,
for
instance,
test
Linux,
okay,
yeah.
D
So
the
the
experience
that
I
had
was
like
I
was
just
getting
frustrated.
I,
don't
know
how
many
times
you
all
have
tried
to
open
the
carrying
the
global
mind.
Job
and
you'll
just
get
a
white
page
staring
you
in
the
face
for
30
seconds
and
then
eventually
you'll
get
a
504,
and
then
it
will
maybe
load,
and
so
I
got
like
a
hint
from
looking
at
the
v8
test
suite
because
they
output
j
unit
by
default-
and
I
found
like
just
going
through
all
the
v8
tests-
was
really
really
fast.
D
B
B
Yeah
sorry
so
I've
done
some
refactoring
of
the
Fedora
posts,
sorry
office,
restarting
fedora,
20
for
removing
21
and
20,
but
on
suggestion
from
rod,
I
actually
took
a
look
at
how
much
we
could
share
and
it's
actually
a
lot.
So
I
still
have
somewhere
to
do
just
some
minor
stuff,
but
hopefully
we'll
end
up
this
week.
And
that
means
we
just
have
one
job
for
all.
Fedora
hosts
I've
also
added
a
bit
of
redundancy,
especially
with
Laura
at
Rackspace.
B
We
they
have
generously
provided
us
with
another
computing
host
in
Sydney,
which
means
that
we
now
have
US
Europe
and
Sydney
with
equal
amounts
of
quota
which
is
like
128,
gives
you
RAM
and
pretty
generous
amount.
So
generous
went
to
very
generous,
which
is
great.
So
what
I'll
do
is
start
moving
some
hosts
over
from
digital
ocean
to
there
just
to
have
a
bit
more
quota
on
digital,
seeing
how
we're
closing
in
on
our
quota
limit
there
and
finally,
broad
invited
me
into
this
Jenkins
to
a
playground
at
our
release.
Thank
you,
so
I'll.
B
A
Look
I'll
carry
on
from
that
and
just
say
that
them,
the
Jenkins,
do
you
think
was
simply
that
I
did
a
an
upgrade
on
this
CI
release
machine
and
we
don't
from
Jenkins
1.5
to
two
point:
seven
and
it's
sure,
Johan
discovered
that
it
was
because
they
had
marked
2.7
as
a
new
LTS
line
and
suddenly
just
put
that
into
their
apt,
Reaper
pepper,
and
so
we
got
the
upgrade
automatically.
It
seems,
though,
that
they
that
you've
opted
into
a
explicit
upgrade
of
the
database.
B
That
up
great
thing,
some,
like
any
of
us,
did
I'm
not
sure
who
did
but
I
don't
see
that
button
and
longer.
So
it's
been
pressed,
okay,
something
okay,.
A
Well,
with
with
fully
in
with
middle,
so
it
turns
out,
it
didn't
really
break
anything.
It
was
nothing
to
fix,
except
for
the
fame.
You
know
I
went
in
that
file.
You
went
to
so
they
say.
I
relate
machine
is
running,
Jacob,
stupid
several
email,
just
don't
yes,
and
we
might
want
to
consider
upgrading
CI,
sir
Anish,
given
that
its
own,
it
seems
to
be
good.
A
You
go
ahead,
yeah,
that's
mainly
because
of
a
theme
it
was
applied,
so
it's
not
include
Jenkins
to
it's,
because
it's
been
thing,
although
I
believe
drink
industry
makes
that
easy,
I
gohan
also
suggested
we
could
get
the
website
folks
in
to
help
us
to
theme.
One
specifically
for
know:
I
know
whether
they
it's
glamorous
enough
work.
Try
not
to
do
that.
We
could
try.
Aside
from
that,
I
think
I've
got
anything
meaty
to
report
premised.
A
You're
not
Lisa
working
again
now,
if
we
sort
of
needed
a
better
way
of
knowing
when
nightlies
not
working,
because
it's
not
the
first
time
that
not
be
stopped
and
so
I'm
thinking
about
that,
but
in
at
least
started
again
because
of
the
cause
stuff
enabled
on
Jenkins
and
that
stopped
the
trigger
from
working.
So
I
had
to
cut
up
the
some
new
cool
stuff
to
get
around
them:
okay,
who's!
Next.
C
I'm
next
in
the
list,
so
I'll
go
so
just
as
you
I
mentioned
general
work
to
keep
the
pp's
machines
working.
I
added
s390
to
the
v8,
nightly
test
jobs
for
master
plus
and
did
the
same
thing
for
the
regular
regression
job.
Since
you
know,
once
we
moved
up
to
a
new
if
level
of
the
eight,
so
there
seemed
to
be
running
without
any
problems.
It's
currently.
If
you
launch
them
against
an
earlier
version.
I,
you
know
created
the
job,
so
they'll,
basically
just
stick.
The
best
report
is
it
reported
his
past.
C
Now,
once
we
actually
show
up,
oops
and
I
will
start
to
configure
them
and
I'm
currently
adding
some
test
machines
from
a
new
OpenStack
environment,
OSU
OSL,
it's
a
new
node
and
newer
versions
of
the
software
which
we're
hoping
will
make
things
more
stable
and
it's
basically
at
the
you
know,
they've
configured
a
test
one,
so
we
can
add
some
machines
validate
that
they
work
and
there's
some
other
teams
that
use
the
same
environment.
We're
going
to
do
the
same.
C
E
Yes,
I
just
went
to
visit
the
new
Jenkins
to
and
found
out
that
TMS
I
is
broken
in
Windows.
Oh
that's!
Next!
For
me,
I
haven't
done
much
in
bills.
I've
been
monitoring
the
art
arrive,
the
shared
are
driving
that
the
raspberries
used,
because
the
new
test
binary
harm
job
that
I
did
was
feeling
it
up.
I'm
still
looking
for
solutions
because
using
it
clean
takes
five
minutes
out
of
each
job
and
it's
not
needed
I'm
thinking
of
just
checking
the
other
half
space
and
running
it.
If,
if
it's
too
low
miles.
D
Inside
of
the
notes,
I've
included
a
link
to
the
pull
request
for
that
reporter
could
use
some
more
eyes
on
it
once
we
have
signed
off
on
that
and
that
lands
I'm
actually
going
to
be
replacing
the
current
canary
in
the
gold
mine
runner
that
we
have
with
using
the
JJ
unit
by
default,
I've
been
using
it
in
the
release
candidate
and
for
other
releases.
There
was
some
weird
stuff
with
XML,
but
that's
may
be
sorted
now.
I
just.
C
D
This
would
not
change
canary
in
the
gold
mine
at
all.
The
j
unit
reporter
is
behind
a
new
flag.
If
I
end
up
actually
changing
anything
to
do
with
flags.
Oh,
that
will
be
a
minor
that
will
be
a
major
bump
rather
than
a
minor
which
is
set
up
for
right.
Now.
There
was
a
review
that
was
standing
there.
There
might
be
a
better
interface,
but
what
would
be
changing
would
be
our
CI
job
to
use
the
new
reporter
instead
of
the
old
room.
C
D
Know
the
reporters
are
all
like.
Basically,
you
can
give
a
different
flag
for
a
different
reporter
and
canary
in
the
gold
mine.
So
it's
just
a
new
flag
but
yeah
the
advice
was
maybe
putting
all
the
reporters
behind
a
specific
reporter
flag,
but
that
would
come
in
a
major
release,
not
in
a
minor,
then
still
not
a
ton
of
work,
but
just
did
a
little
bit
more
of
playing
around
with
the
paralyzed
version,
the
canary
in
the
gold
mine,
there's
just
some
weird
stuff
in
there
that
I
haven't
been
able
to
quit.
B
D
D
Sorry
about
that,
my
mic
unplugged
last
thing-
I'll
be
really
quick,
is
just
started
thinking
about
how
we
could
use
canary
in
the
gold
mine
for
testing
changes
to
binary,
specifically
like
making
sure
that
there
aren't
a
bi
changes
in
between
different
versions
or
specific
v8
changes,
and
then
Johan
I
may
want
to
talk
to
you
offline
about
some
of
the
stuff
that
I'm
thinking
about.
For
that
sounds
good
and
that's
it
for
me.
B
Okay,
so
that's
the
roll
call,
so
I
guess
we're
moving
to
the
agenda
of
today.
The
first
issue
would
be
inviting
Philip
J
to
the
bill
group
and
I'll.
Just
do
a
quick
background,
so
Philip
is
contributed
to
both
node
core.
B
C
E
B
Sweet
so
welcome
Philippe's
on
vacation
right
now,
but
he's
probably
having
a
beer
somewhere
and
celebrating.
So
that's
called
congratulations.
Indeed.
I
can
take
care
of
onboarding
that
the
plan
was
to
do
any
Mon
boarding
with
imuran
while
ago,
but
that
never
happened
so
I'll
try
and
do
it
this
time
and
kind
of
figure
out
what
what
the
baseline
of
knowledge
we
need
to
spread
to
people
joining
the
project
is
and
then
I'll
try
to
put
that
into
some
kind
of
guide
we
can
have
as
part
of
the
bill
group.
E
B
A
A
Would
like
to
be
our
ship
images
with
it,
but
the
workflow
for
shipping
images
going
to
download
binaries
off
our
website
and
use
those
rather
built
right.
Have
them
build
them
themselves
in
them
in
contains
because
it's
a
waste
space
and
makes
and
takes
too
much
time
and
it's
not
ideals
website,
but
we
can't
they
can't
use
the
binaries
we
ship
now
because
they're
that
they
need
to
have
the
H
we
compiled
against
musim
wrong.
A
I
I've
been
pushing
back
on
shipping
Musil
wineries
in
there
for
this,
because
it's
like
once
we
start
doing
it,
for
you
know,
x64
we're
going
to
be
a
spur
to
do
it
on
for
other
architectures,
and
then
these
are
the
things
that
could
be
asked
to
happen.
Well,
it's
just.
It
just
seems
like
a
slippery
slope
to
adding
lots
more
binaries,
so
I'm
not
I'm,
not
sure
we
can
come
to
a
conclusion.
You,
because
I
haven't
seen
any
great
suggestions
in
that
thread.
B
As
well
so
I've
also
been
pushing
back
like
making
muscles
support
with
semi-official
efficient
I
just
think
the
disability.
You've
added
is
great,
and
you
know
if
we
can
keep
it
like
this
for
a
couple
of
months
and
we've
had
like
packages
from
from
alpine
Linux
actually
contribute
and
help
us
improve
this
as
sweet
and
coverage.
So
you
know
I
would
say
this
is
a
step
in
the
right
direction,
but
there's
still
you
know
maturity
that
needs
to
be
kind
of
grown
there.
A
Might
be
a
alternative
way,
we
could
do
it
me,
maybe
making
the
Pine
Ridge,
but
not
for
me,
location,
I,
don't
know
if
that
isn't
too
much
the
same
as
well.
C
A
That
we
do
that
without
okay
alpine
alpine
machine
in
jenkins.
Now
this
is
about
binary
it's
because
they
want
to
be
able
to
ship,
but
the-
and
this
is
actually
a
fairly
big
deal,
because
our
pine
is
very
popular
forum
for
dr.
users
and
dr.
becoming
more
popular
and
and
they'd
love
to
be
able
to,
but
they
current
passes
to
build
them
binary
themselves,
and
that's
not
ideal,
so
it'd
be
good.
If
we
could
find
a
way
to
support
them,
I
don't
want
to
push
back
just
them
and
let
that
be
it
I.
B
A
Actually
came
from
a
quest
from
from
feud
or
about
adding
ll
node
to
Jenkins
so
that
it
could
have
just
covered
and
it's
I,
it's
becoming
a
fairly
important
project.
We
had
brain
Greg,
covering
at
the
other
day
as
well.
So
it's
it's.
It's
getting
more
visibility
and
could
be
a
very
important
tool
for
notes
future,
but
it's
not
part
of
our
all.
Yet
it
maybe.
A
C
A
A
A
In
there
1
stage,
2
helps
you
import,
but
as
other
projects
come
and
ask
for
assistance,
how
do
we
qualify
them?
And
so
the
there's
multiple
concerns
here,
there's
a
concern
about
how
much
capacity
we
have
and
your
hands
preferring
up
the
best
me
we've
been
getting
more
from
various
sources.
Sequestering
may
not
be
a
problem
right
now,
but
the
other
thing
is
the
intention
of
the
donors,
so
when
they
donate
hardware
2
to
node.
A
Are
they
okay
with
us
using
it
for
non
core
stuff,
and
do
we
risk
putting
our
donors
offside
by
diluting
that?
So
that's
this.
The
main
two
concerns
I
have
there
and
you
know
if
we
started
getting
access
to
some,
then
we'll
get
more
and
then
we
get
up
to
deal
with
security
concerns
for
having
all
these
people
that
we
don't
know
accessing
out
by
where
yeah.
A
A
D
I
had
had
a
conversation
with
Jordan
just
trying
to
figure
out.
You
know
specifically,
it
started
with.
How
can
we
get
support
for
installing
our
seas
and
you
know
other
test
releases
via
nvm,
and
so
he
is
working
on
a
one
point,
0
version
of
nvm,
which
will
be
able
to
do
that,
but
the
biggest
thing
that
he
he
doesn't
have
working
right
now
is
the
ability
to
test
nvm
compiling
from
source.
D
He
doesn't
have
the
architecture.
That's
able
to
do
that.
It
has
something
to
do
with
timeouts
on
untraveled.
Just
doesn't
give
him
enough
time
to
actually
do
a
build
before
canceling
everything,
so
we
had
discussed
potentially
using
the
node
infrastructure
to
run
the
nvm
test
suite
on
the
various
you
know
in
Fisher
the
architectures
that
we
deploy
to.
He
seemed
really
interested
in
it.
I
had
started
playing
with
it
a
bit,
but
then
like
ran
into
some
roadblocks
with
the
way
the
nvm
test
suite
was
set
up.
D
I
have
not
revisited
it
since
this
was
maybe
about
like
three
weeks
ago,
but
it's
definitely
something
I
can
dig
into
more
and
something
that
he
was
interested
in
in
having
in
general,
I
think
for
these
tools.
It
makes
a
lot
of
sense
to
be
running
the
tests
on
the
architecture
we
ship
to
so
I'd
like
to
see
that
happen
and
I'd
also
really
like
to
see
RC,
support
and
nvm,
which
you
know
this
testing
is
a
blocker
for
that.
C
A
C
A
B
C
B
Yeah
I
mean
if
this
was
just
testing
on
like
different
flavors
of
Linux
or
something
we
call.
We
spin
up.
New
hosts
on
demand,
I
mean
that's
pretty
low
key
for
us.
I
wouldn't
have
a
problem
maintaining
that
if
someone
needs
a
host
for
24
hours
every
now
and
then
but
I
just
think
if
or
to
deploy
something
like
throughout
our
infrastructure.
That's
it's
a
bit
trickier.
D
Yeah
this
was
also
an
interesting
point
that
had
come
up
was
at
least
you
know
the
nvm
project.
I
think
this
is
the
case
for
many
projects
in
the
node
ecosystem.
Kind
of
treat,
CI
is
something
that
happens
on
every
commit,
because
you
know
providers
like
Travis
or
so
sandboxed
that
they
can
do
that,
and
it's
just
not
something
that
we're
doing
or
necessarily
want
to
do
so.
That
I,
think
is
a
big
is
a
big
thing
for
us
to
figure
out
is
like.
D
If
people
want
to
come
and
be
using
our
infrastructure,
are
they
looking
for
an
automated
CI,
or
are
they
going
to
be
okay
with
something
that's
not
run
for
every
commit
on
every
PR
or,
alternatively,
are
we
open
to
offering
other
projects
if
they're
more
sandbox
the
ability
to
you
know?
Have
it
automated.
C
Yeah,
that
is
a
good
point
that
there's
different
levels
of
you
know
it's
one
level
to
replace
their
CPC
I,
but
it's
a
different
thing
to
say:
well,
it's
not
going
to
replace
their
CI,
but
we're
going
to
have
our
cai.
Do
some
testing
on
you
know
they're
on
those
extra
packages
you
know
possibly
because
we
have
more
breath
or
whatever,
but
not
necessarily,
you
know,
replace
and
run
every
time
they
would
run
so.
D
D
A
lot
of
sense
is
maybe
it's
not
a
CI
that
he
uses
regularly
as
part
of
the
project,
but
it's
something
that
can
be
run
against
before
a
new
nvm
release
to
make
sure
you
know
in
like
the
same
environment
that
are
like
mirrored
from
the
build
of
our
builds,
that
there
aren't
new
regressions
before
doing
a
release,
and
that
could
be
an
interesting
way
of
approaching
it
to
different
projects
of
like
hey.
You
know
like
we
have
this
infrastructure.
A
I'm
not
sure
that
really
solves
any
security
concerns
them
because
it
like
within
vm
sure
we
know
jordan
and
we
can.
I
think
we
can
be
pretty
confident
there
comes
with
me-
comes
with
Airbnb
and
tasty
39.
So
he's
you
know,
he's
got
some.
It's
got
things
to
lose
your
fears
to
do
things
wrong,
so
I
think
yeah
and
same
with
Fedor.
We
can.
We
can
deal
with
that.
We
might
be
able
to
do
with
report
as
well.
A
D
C
A
See
that
is
good,
but
it's
simply
giving
them
access
to
machines
that
you
know
we
do
everything
else
on
and
and
and
we
have
keys
and
other
secrets
that
maybe
label
from
CI
we
have
the
ability
to
possibly
mess
with
our
CI
test
runs.
You
know,
there's
all
these
little
security
problems
would.
Rather
you
know
that
that's
what
travelers
can
work
trusting
others
can
work
is
because
they
do
either
containers
or
end
or
temporary
infrastructure.
So
they
can
keep
things
fairly
clean,
but
we
don't
do
that.
We
don't
have
great
ways
of
doing
that.
Mm-Hmm.
A
And
Arab
donors-
okay
with
that,
but
we're
running
machines
on
your
infrastructure
that
we're
not
even
testing
going
on
mm-hmm
and
to
be
honest,
I,
don't
even
want
to
approach
the
digital
ocean
or
Rackspace
about
this,
because
the
more
we
have
to
have
all
good
conversations
with
them
about
node,
the
more
likely
we
are
to
run
into
a
situation
where
they
say
all
in
dance
for
any
money.
No
here
visit
that.
A
D
A
E
E
B
I
mean
there's
also
the
fact
that,
for
instance,
nvm
wanted
access
to
architectures
it
couldn't
test
on.
So
you
know
this
would
like
12
into
PPC
and
freebies,
and
that
makes
it
even
harder.
I
mean
if
we
have,
let's
say
I
mean
ansible,
has
support
for
spawning
images
and
Rackspace
or
whatever
like
it
would
be
pretty
easy
for
us
to
someone
thinks
us.
We
spin
up
something
that's
available
for
24
hours,
I
mean
that
would
work
well,
but
like
scaling
it
on
to
all
our
types
of
architectures.
That's
really
the
tricky
part
here
so
yeah.
B
C
A
C
I'm,
certainly
interested
you
know,
from
the
perspective
of
it's
difficult
from
people
to
build
across
all
the
architectures,
so
if
this
is
a
way
to
get
access
for
people,
you
know
for
projects
that
are
important
to
know
in
the
ecosystem.
It's
it's
more
realistic
than
each
of
them
building
out
their
own
infrastructure
writing,
including
power
and
a
ringing.
You
know.
B
B
Think
it's
six
weeks
now,
since
we
last
spoke,
my
suggestion
was
to
just
try
it
do
a
quick
catch-up
and,
like
you
know,
figure
out
where
were
their
point
of
differences
where
so
there
was
a
few
things
you
race
during
that
call
in
terms
of
security,
access
and
so
on,
so
that
still
needs
to
be
wrangled
out
and
that's
pretty
much
where
we
are.
Hopefully
we
can
resume
that
work
pretty
shortly
and
just
make
progress,
but
I
wish
I
could
say
more,
but
that's
pretty
much
where
we
are
at
the
moment.
A
B
That's
correct,
so
the
800
bucks,
ish
or
whatever
it
was,
is
for
like
three
Mac
minutes,
16
gigs
around
which
you
gives
us
like
I,
think
it's
one
terabyte
perd
mini,
which
gives
us
like
a
lot
of
leeway
for
hurting
whatever
VMs
we
need,
and
the
beauty
was
that
they
also
had
VMs
already
like
a
different
versions
of
OS
X,
already
available
through
there
VMware,
which
means
exemplar
for
us
to
deploy
it.
Yeah,
brilliant.
B
B
Okay,
cool
well
bola
will
push
it
forward,
as
we
do
I'm
good
with
that.
So
unless
anyone
else
wants
to
bring
something
up,
let's
move
on
so
next
issue
would
be
the
arson
can
point
to
mirror
and
releases,
which
is
the
255
and
I've
already
brought
most
of
that
up
in
my
stand-up.
So
we're
pretty
much
there
and
what
remains
to
be
done
is
just
to
deploy
a
script
on
our
main
host
noor
jesser
org
that
pushes
to
our
sink
host
on
deployment
and
throw
a
con
job.
B
And
besides
that,
we're
pretty
much
done
unless
we
should
have
another
round
of
discussion
with
regards
to
end
of
life.
But
the
problem
right
now
seems
to
be
that
there
is
no
good
alternative
for
our
sink
in
terms
of
simplicity,
of
use
automation,
but
with
the
encryption.
So
that's
pretty
much
where
we
stuck
at
the
moment.
B
So
yeah
the
moving
off
of
that
then
finally
run
and
deployed
in
ojs
get
up
but
issue
404,
which
is
what
Phillip
is
missing,
but
just
to
summarize
the
issue
it
is
about.
How
do
we
deploy
the
github?
But
how
do
we
handle
the
shared
secrets
because
it
needs
access
to
the
main,
get
up
organization
and
will
likely
also
need
elevator
access
to
Jenkins,
Jenkins,
CI
or
CI
release,
and
perhaps
other
endpoints,
which
makes
the
security
issue
more
important.
A
B
A
C
Just
I
just
had
one
to
ask
one
question:
I
may
have
asked
this
before
and
forgotten
the
answer,
but
I'll
ask
again
anyway
in
terms
of
the
AIX
machines,
it
would
be
nice
if
there
was
somewhere.
I
could
store
binaries
that
ansible
could
pull
from,
because
we
don't
have
the
same,
no
easy
access
to
external
repos,
where
we
can
just
pull
things
down
and
install
them
and
I'm.
A
A
They
pull
the
slave
jar
from
there.
So
there's
already
downloading
it's
a
vibration
there,
the
other
any
of
you
used
is
the
the
OSX
machines
currently
use
that
machine
as
a
an
ssh
tunnel
for
an
HTTP
proxy
which
runs
on
that
machine
coherence
and
network
issues
we
have
with
voxer.
So
we
could
reuse
that
machine
to
serve
binaries.
If,
if
you
need
to
like
what
you
just
pick,
Lee
I
think
we're
running
already
running
engine
X
on
there,
so
you
could
pull
a
ok
another
pass.
Yeah.
C
There
are
things
that
I,
basically
like
there's
a
site
called
bull
freeware,
where
you
can
go
down
and
pull
down
a
RX
binaries
for
different
releases
from
there,
though
you
have
to
like
prove
you're,
not
a
robot
kind
of
thing
wrong.
So
I
can't
just
script
that
easily
so
I,
don't
think
there
anything
that
you
know
we
need
to
protect
people
from
getting
at
necessarily.
We
don't
necessarily
want
to
be
a
distributor
of
them.
You
know.
Ideally
it
would
be
something
that's
available
internal
to
our
CI.
C
B
We
do
that
or
just
spin
up
another
machine.
We
have
room
I,
don't
have
a
strong
opinion,
either
works
for
me.
I
guess.
C
C
C
A
C
B
A
I
did
wonder
whether
I'm
it
you
know
since
ring
we've
got
constraints
with
people
availability,
whether
we
should
consider
a
less
frequent
meeting
time,
or
is
this
one
too
infrequent
as
it
is,
or
you
want
to
discuss
that
I
was
thinking
my
own
evil,
exes
here
and
also
myself,
but
maybe
Alexis
just
can't
afford
to
come
in
general.
We.