►
From YouTube: Node.js User Feedback Initiative Meeting
Description
B
All
right
welcome
everybody
to
the
end
user
feedback,
meaning
today
we're
gonna
be
focusing
focusing
a
little
bit
on
the
enterprise
use
cases.
There
are
a
few
folks
here
who
are
deeply
in
the
enterprise
world,
so
perhaps
we
can
just
kick
it
off
by
a
quick
intro
from
everybody,
and
then
we
can
talk
about
the
agenda
and
some
of
the
topics
we
would
like
to
go
through
so
I'll
start.
B
My
name
is
Ivan
Austria
I'm,
the
personal
architect
that
tell
us
tell
us,
is
a
telecom
company
in
Canada
and
we're
heavily
invested
in
the
nodejs
ecosystem
and
we're
looking
for
ways
of
how
we
can
provide
some
feedback
and
or
value
back
to
the
ecosystem
in
any
shape
or
form.
Given
our
deep
investment
in
it
and
our
dependency
on
it.
C
My
name
is
Joe
Cheng
I'm,
the
principal
architect
of
customer
engineering,
web
platform
at
Walmart
labs,
so
warmer
and
warmer
labs
has
been
a
longtime
user
of
no
J's.
You
know
because
warm
is
such
a
big
company.
We
have
many
many
teams
whose
use
no
J's,
that
we
have
teams
that
we
don't
even
know
about
that
has
been
using
noses
for
some
something,
and
in
the
past
three
years
we
actually
migrate.
The
mamata
comm
website
to
node.js
and
we've
been
heavily
invested
and
migrating
more
and
more
things
to
node.js.
C
D
Sure
hello,
my
name
is
Mihaly
Petrus,
no
I'm
a
project
manager
acting
in
Portland
area,
which
PMI
Institute
and
Romanian
American
Society
and
right
now,
I'm
a
user
feedback.
Member
Thank
You.
A
Try
to
be
brief,
this
is
done
quick,
so
dance
off,
I
created
the
the
nagas
user
feedback
initiative
and
I
champion
this
initiative
to
the
community
committee
of
no
dis
project.
I
am
a
long,
long
time,
enterprise
sort
of
connective
tissue.
You
know
starting
to
startups,
to
note
firm
and
note
source
that
are
dedicated
to
you,
know,
kind
of
serving
that
intersection
between
noches
and
the
enterprise.
B
So
this
is
hopefully
the
start
of
a
open
dialogue
that
can
help
feed
some
of
those
conversations
in
both
ways,
so
that
you
know
when,
when
big
companies
that
say
like
ours
are
rolling
into
new
versions
of
node
and
seeing
perhaps
issues
or
improvements
or
values
being
created
from
certain
features
or
certain
investment
and
features
that
yields
multi-million
dollar
kind
of
business.
But
more
importantly,
is
serving
millions
of
users.
B
It's
a
that
context
has
been
missing
and
hopefully
we
can
surface
that
so
like
I'm,
just
gonna
piggyback
off
what
Joe
said.
We
also
did
a
recent
update
to
our
website
where
we
went
entirely
nodejs
in
the
past
couple
of
years
and
that's
you
know
we
are
talkin
Canada
we're
serving
millions
and
millions
of
customers
and
it's
very
core
and
integral
to
our
business,
the
health
of
the
ecosystem
and
the
health
of
our
applications
and
since
they're
Atari
nodejs,
based
having
that
connection
to
the
community
and
having
that
value
conversation
surface
is
very
important.
B
So
just
a
structured
conversation,
a
little
better
I
want
to
understand
a
little
bit
more
from
everybody
like
what
does
it
actually
mean
that
you're
using
nodejs,
you
know
again,
like
Joe,
said
your
website
isn't
no
js',
but
what
does
that
actually
mean?
What
are
you
actually
using?
What
is
the
ecosystem
for
your
development?
Tooling,
look
like,
and
is
it
mostly
JavaScript
with
nodejs
as
an
accelerator
for
tooling,
or
is
it
you
relying
on
the
GS
itself
for
running
your
backends
and
your
API
is
in
your
systems.
C
I,
take
that
one
know,
but
before
nodejs
and
warm
our
lab
is
a
very
heavy
Java
user
and
as
up
to
date,
we
are
still
very
heavy.
Jvm
user
I
mean
not
strictly
the
Java
language
itself.
For
example,
we
are
work
teams
who
are
using
closure
and
we
also
have
teams
who
are
using
go
lengths
for
like
high
traffic
routing
and
data
routing
applications
on
the
front
end
of
the
website
itself.
We
we
were
using
Java
and
spring
web
flow
in
Java
and
a
bunch
of
a
different
style
script
framework
like
backbone
and
jQuery.
C
So
when
we
moved
to
no
js',
we
were
primary
thinking
of
using
node.js
to
back
to
power,
react
web
application.
So,
and
that
means
that
means
that
no
js'
has
replaced
all
of
the
web.
Brennan
and
powering
would
react
to
run
the
website
and
in
the
backend
as
an
orchestration
layer
to
connect
to
all
the
backend.
C
So
the
backend
is
still
based
on
Java
and
a
lot
of
our
tooling
existed
as
a
Java
framework
like
we
use
a
nexus
with
Jenkins
for
build
and
where
we
are
building
new
tools
in
Java
that
that
are
complementing
the
existing
Nexus
and
the
Jenkins
ecosystem.
For
example,
we
actually
have
teams
who
built
an
NPM
server
on
top
of
Nexus,
and
we
have
teams
will
be
a
build
system
on
top
Jenkins.
So
no
JSF
now
is
mainly
used
as
an
orchestration
for
for
the
web
applications.
C
B
Middleware
for
the
most
part,
yes,
okay
and
and
how
you
know,
how
reliant
are
you
on
that
middleware
in
terms
of
your
business
operations,
like
just
to
quantify
that
a
little
bit
more
and
what's
the
what's,
the
opportunity
cost
for
you
to
even
consider
trying
something
other
than
node.js?
Is
that
something
that
you're
heavily
invested
in
in
terms
of
that
metal
or
technology
being
reliant
on
ojs,
given
the
front-end
dependency,
or
is
that
just
a
investment
that
you're
doing
now?
But
you
know
you
don't
have
a
dependency.
Oh
I
will.
C
Say
we
are
deeply
integrated
with
nodejs
as
far
as
running
our
website,
I
mean
we
were.
You
know
like
within
the
first
year
of
deciding
to
convert
in
ojs,
we
are
convert.
Almost
every
team
we
have.
A
few
hundred
developers
were
working
on
a
ecommerce
site
itself
and
within
the
first
year
we
convert
all
the
apps
like
we
have.
C
We
have
some
couple
major
app
like
we
have
the
home
page
when
you,
when
you
get
warm
at
home,
that
that
home
page
and
every
search
we
have
our
item
page
category
and
basically
everything
that
has
to
it.
Wilma
comm,
we've
converted
our
team
to
to
no
js'
in
a
friendly
and
orchestration
in
the
middle
layer.
So.
C
C
So
we
thought
that
okay,
everything
is
JavaScript
so
now,
while
developers
can
focus
on
one
thing
and
that
that
has
definite
advantage,
because
we
we
were
able
to
get
our
developers
more
developers
have
to
work
on
wider
area
of
things,
but
at
the
same
time,
because
we
move
into
an
odious
and
obviously
notice
is
more
than
JavaScript
and
underneath
and
nodejs
the
way
notice
work
at
the
asynchronous
nature
as
a
as
a
very
is
very,
very
new
to
a
lot
of
our
developers.
And
so
training
is
a
has
been
our
biggest
thing.
C
Me
personally,
I
I
I
mean
obviously
I
personally
I
don't
find
documentation
lacking
and,
of
course,
I
dig
into
the
code
more
frequently,
I
actually
look
at
the
documentation,
but
for
a
lot
of
our
developers,
the
some
of
our
giver,
who
who
are
like
I,
guess
I'm,
who
need
a
little
bit
more
hand-holding,
then
just
kind
of
go
and
dig
into
the
color
and
figure
out
everything
on
their
own
yeah,
I
I
would
say
that
documentation.
You
know
it's
a
it's
kind
of
fragmented
as
far
as
the
documentation
goes,
I
always
go
to
node.js.
C
There's
cause
documentation
directly
to
look
up
anything
about
node.js
but,
as
you
know,
those
are
kind
of
white
page
reference,
so
you
really
have
to
to
read
between
a
line
to
figure
out.
What's
going
on
and
and
I
love
all
the
above
developers,
you
know
you
know
a
typical
task
for
them
and
they
just
want
to
to
find
out
okay.
How
do
you
do
this
thing?
What
is
there
any
example
that
that's
one
of
the
question
we
I
get
asked
the
most?
B
Think
the
the
key
takeaway
there
is
that
the
documentation
today,
at
least
on
the
core
side
of
things,
is
very
reference
oriented
as
opposed
to
guides
and
an
example
oriented.
That's
that's
a
good
kind
of
takeaway
Adrienne
I
want
to
switch
to
you
quickly
in
the
same
kind
of
context
of
the
questions
like
what
sort
of
workloads
are
you
using
nodejs
for
and
what
has
been
your
success
factors
and
challenges
and
getting
your
teams
to
adopt
and
use.
E
So
nodejs
is
picking
up
a
lot
in
that
back-end
as
a
front-end
kind
of
style,
with
a
lot
of
the
clients
and
the
teams
that
I'm
working
on
and
the
reason
again,
the
the
the
companies
I
work
with
are
heavy
into
the
Java
ecosystem.
They
have
existing.
You
know
large
applications
built
in
Java
that
they
want
to
interface
with
or
do
something
quickly
with
so
building
these
backing
as
a
front-end
and
their
channels
layer
to
do
specific
things
to
expose
specific
API
is
internally
or
externally.
E
They
do
that
with
NPM
registries.
They
are
self
hosted
on
their
own
artifactory,
Nexus
repository
and
then
the
consumer
of
those
libraries
or
API
libraries
that
are
being
developed
standards
are
consumed.
You
know
your
regular
package
to
use
and
file.
The
problem
is
when
you
have
multiple,
posit
or
ease
you
know:
you'd
set
up
an
AP,
MRC
type
thing
and
name
your
specific
repositories
and
then
explicitly
call
the
the
libraries,
by
name
in
your
package,
JSON
file.
That
creates
a
lot
of
headache
when
we
try
to
do
things
like.
E
It's
a
lot
for
that
part
of
the
organization
need
to
deal
with
and
there's
optimizations.
Obviously
you
can
make
a
process
and
things
like
that,
but
this
is
the
way
it
is
right
now
and
so
where,
if
you
take
maven
as
an
example
of
the
tool,
life
cycles
and
multiple
repositories
and
having
the
tool
actually
be
able
to
query
multiple
registries,
that
kind
of
experience,
I
think
would
be
like
much
better
for
these
types
of
organizations
and
yeah
I,
think
yeah,
I
think.
E
B
For
context,
my
team
right
now
is
about
450
humans
out
of
those
at
least
one-third
our
developers,
the
rest
are
between
QA
design
and
kind
of
project
and
business.
Our
leadership
teams,
so
over
150
to
200
people
are
directly
working
on
code
bases
and,
as
far
a
platoon
comes,
we
have
I,
think
800
individual
applications
and
about
100
or
so
libraries
that
we've
created
and
maintained
internally
Joel.
What's
your
kind
of
stats
look
like
for
that
context,.
C
Sorry
we
yeah
so
just
for
walmart.com
itself.
We
have
I,
don't
have
exact
number,
but
just
for
math
outcome
itself.
We
have
probably
about
300
developers,
it's
just
the
so
called
web
developers
and
that
that
doesn't
include
all
the
you
know,
other
sub
supporting
teams
and
for
teams
we
have
about
10
12
teams.
Each
team
is
anywhere
is
between
10
to
15
developers
give
or
take
so
that
out.
C
B
B
So
this
is
why
that's
important
in
the
context
of
what
you
mention
of
onboarding
and
training
and
knowledge
management,
and
what
they
do
is
mentioning
in
terms
of
complexity
of
sharing
code
and
and
managing
access
to
code
and
and
and
consistency
in
development.
Sorry
in
delivery,
because
it
small
you
know,
I'm
a
factor
of
our
ten
to
a
thousand
seems
of
complexity
when
you
have
these
these
amount
of
teams.
So
anybody
else
on
the
call
has
any
additional
context
side
here,
so
some
people
join.
It
seems
that
they
also
disconnected
may
be
enough.
B
B
I
think
the
the
the
relevant
next
question
to
ask
in
that
context
is,
you
know,
just
to
use
their
recent
examples.
Any
recent
releases
or
changes
or
updates
to
API
is
in
the
core
of
node
that
have
directly
impacted
your
team
or
are
directly
in
your
line
of
sight
like,
for
example,
we
just
we
haven't
a
new,
really
any
security
patch
release
just
today,
I
think
or
yesterday.
Do
you
look
at
those
every
day?
Are
those
impactful
to
your?
You
know,
velocity
of
delivery.
B
E
Sure
so
I
mean
things
move
pretty
slow
in
financial
services,
insurance
companies-
and
there
are,
you-
know,
specific
parts
of
the
organization
that
look
after
that
now
we
could
talk
about
things
like
containers
and
and
passes
kubernetes
to
kind
of
solve
these
issues
and
shift
those
responsibilities
a
bit
a
bit
left
to
the
development
development
teams-
we're
not
there
yet,
so
it
is
slow,
moving
so
being
able
to
upgrade
and
get
the
latest
patches
and
things
like
that.
It
is
a
bit
of
a
process.
B
A
E
The
way
it
would
typically
work
is
you
have
this
security
part
of
the
organization
or
the
risk
side,
monitoring
those
type
of
releases
and
changes
and
giving
or
even
an
audit
right.
Let's
say
an
external
auditor
comes
in
and
says
you
gotta
upgrade
that's
another,
that's
another
event:
I
could
trigger
something
and
then
there's
specifically
a
plan
or
rollout
to
make
the
upgrades
across
the
organization.
E
B
Joel,
do
you
have
any
context
in
terms
of
what
degree
are
you
investing
in
or
looking
at
what
changes
are
happening
in
the
release
cycles
of
node
and
the
kind
of
Saylor
libraries
in
the
ecosystem
around
it?
Like
you,
do
you
actively
look
at
those?
Are
you
actively
looking
for
security
patches
and
updates?
Do
you
review
every
changes
to
API
features
or
changes
in
the
api's
themselves?
Yeah.
C
We
we
have
many
many
people
were
passively
and
actively
monitoring
the
changes
in
the
core
for
our
practice.
We
actually
lock
to
a
version
and-
and
we
don't
update
that
version
unless
there's
some
something
that's
measuring.
You
know
like
we
do
a
we.
We
do
an
update
every
six
month
and
then,
if
there's
any
security
or
it
released,
then
we
would
consider
upgrading.
Otherwise
we
will
act,
a
version
for
about
six
month
period
and,
like
I
said
we
have
developer,
who
are
constantly
monitoring
and
I
mean
all
those
on
our
internal
slack.
C
B
It
sounds
like
and
I
would
echo
the
same
thing
for
our
organization.
A
lot
of
the
interactivity
is
very
reactive.
You
know
we
hear
about
something
on
Twitter
or
we
see
it
in
an
announcement
and
then
either
we're
like
okay,
we
can
upgrade
or
you
know
it's
changed
to
an
API
or
a
core
feature
and
we're
like
okay,
we
don't
use
that
or
we
use
that
might
or
might
not
affect
us.
Is
there
an
opportunity
for
for
that
to
become
a
feedback
cycle
or
a
feedback
loop
like
at?
What
point?
B
Are
you
able
to
invest
something
back
to
the
community
by
either
testing
those
things
ahead
of
time?
Perhaps-
or
you
know,
aside
from
an
individual
level
which
I
know
we
all
do
out
of
passion
but
like
as
an
organization?
How
do
we
drive
that
you
know
feedback
cycle
or
investment
back
into
the
ecosystem
and
just
to
give
a
couple
of
solid
examples?
B
You
know
when,
when
you
know,
there's
LTS
and
there's
release
management
and
all
of
that,
but
there's
no
reason
that
we,
as
development
teams,
can't
go
and
put
all
of
our
development
environments
or
staging
environments
on
no.10
and
start
testing.
It
make
sure
all
of
our
apps
are
working
and
if
any
issues
come
up
correlated
we
can,
then
you
know,
provide
a
direct
feedback
to
the
core
teams.
Now
the
question
of
how
do
you
do
systematically
do
that,
but
I
think
there's
an
opportunity
there
too
or
likewise.
B
B
You
can
see
how
many
downloads
a
certain
package
has,
but
it
doesn't
tell
you
that
its
operating
on,
let's
say
Walmart's
back
ends
and
it's
a
critical
path,
the
path
of
the
delivery
for
Walmart
or
tell
us
or
any
other
financial
system.
So
how
do
we
effectively
are
able
to
surface
those
things
so
that
the
investment
in
either
the
core
or
the
packages
around
the
ecosystem?
Are
you
know
at
least
given
more
attention,
if
not
potentially,
more
funding
even.
C
So
we
do,
we
do
have
a
lot
of
teams
who
are
kind
of
officially
constantly
trying
out.
Whenever
there's
a
new
major
version,
we
don't,
we
don't
go
for
them
to
mine
of
a
patched
version
unless
there's
a
big
something
improvement,
but
whenever
a
new
major
version
comes
out
with
plenty
of
our
developers
and
teams
already
trying
out
and
running
their
apps
on
staging
and
test
and
Raman
t-test.
C
So
this
is
where
I
would
say
that
so
far,
everything
is
that
coming
out
from
no
core,
it's
actually
pretty
good,
because
you
know
we
started
with
version
4
and
when
we,
when
we
migrated
the
6,
8
and
10
every
time,
even
just
a
fist,
the
first
release
of
each
major
we
we
were
able
to
migrate
to
it
fairly
easily.
Without
any
issue,
I
mean
we
just
like
4
6
8.
We
just
kind
of
change
the
version
and
boom
everything
this
works,
and
obviously
you
know
we
take
typically
a
little
bit
into
it.
C
We
find
it's
small
issues
that
we
just
just
fix
so
and
if
we
find
any
big
things,
we
do,
we
don't
go,
go
to
the
core
and
may
file
file
issue
or
ask
about
it
or
we
will
go,
go
around
and
look
for,
look,
look
for
activities
and,
more
often
than
now
any
small
issue
we
found.
We
find
that
entire
internet
who
people
who
used
notice
they
already
they
over
all
or
you
have
the
same
issue
and
they
would
already
be
an
issue
file
or
still
have
been
talking
about
already.
C
F
C
B
F
B
So
to
the
point
of
making
something,
systemic
I
would
imagine,
it's
gonna,
be
a
complex
and
be
costly
to
say,
Adrienne,
hey
how
about
you
go
and
test
the
latest
version
of
10,
no
10
on
all
of
your
application
and
perhaps
run
some
diagnostic
and
send
it
back
to
anonymously
anonymous
anonymous
data.
Hopefully
not
nothing.
Business-Critical,
send
some
feedback
back
to.
Let's
save
the
note
core,
perhaps
on
things
like
performance
does
GUP
grading
from
you
know
ten
point
one
to
ten
point
to
improve
or
decrease
performance
in
any
significant
way
to
your
business.
B
Do
you
see
any
sort
of
regression
issues
that
weren't,
captured
or
wouldn't
addressed
in
the
kind
of
public
forum?
Is
that
something
you
were
able
to
do
at
all?
Is
there
a
cost
constraints
to
that,
given
that
perhaps
there's
a
some
form
of
systematic
way
to
do
that
and
some
form
of
way-
and
maybe
this
is
the
group
to
kind
of
surface
how
to
do
that?
But
let's
say
those
things
exist.
Are
you
able
to
do
it.
E
E
So
I
don't
think,
there's
like
if
there
was
a
systematic
way
to
do
that.
I
think
definitely
the
organization
I
work
with
would
take
advantage
of
that
I
mean
the
more
data.
The
better
and
and
being
able
to
report
that
back
to
the
broader
community
would
definitely
be
a
value,
especially
those
that
are
moving
quicker
to
than
the
others.
E
E
I
think,
as
these
organizations
move
more
towards
automating,
a
lot
of
their
testing
environments
or
just
testing
or
delivery
application
delivery
process
in
general,
those
concerns
tend
to
go
away
and
then,
when
you're
able
to
you
know
create
ephemeral
environments
on
the
fly
to
do
some
upgrade
tests.
Just
some
quick
progressions
again
those
problems
go
away,
I
think
for
sure.
In
the
current
process
there
will
be
added
costs.
Charge
backs
from
you
know,
with
the
QA
perform
his
team
or
something
like
that
back
to
the
team.
B
B
I
cannot
do
that
without
fundamentally
changing
how
my
pipelines
are
built
so
that
I
can
still
do
my
regular
pipelines
and
I
can
run
this
in
parallel
somewhere
else,
because
all
of
my
version,
locking
is
happening
in
docker
files,
I
have
to
figure
out
a
way
to
systematically
change
that
either
on
runtime
of
the
pipeline
or
it's
working,
every
single
repo
and
having
a
different
version
lock
in
it.
But
I
would
love
to
be
able
to
do
it
and
I,
don't
think.
B
F
I
guess
is
a
good
way
to
put
it.
I
I.
I
would
also
echo
that
they
don't
have
that
they
don't
necessarily
have
the
cost
concern,
but
it's
more
of
an
expertise,
concern
in
terms
of
okay
like
or
do
we
want
to
do
this
or
even
being
aware
of
it?
You
know
or
does
this
exist?
Is
this
something
we
would
get?
What?
F
C
Would
mean
that
it
would
we
will
be,
it
will
be
visible
on
the
radar
and
the
management.
Well,
what
could
possibly
ask
about
it,
and,
and
once
we
make
this
official,
it
means
it
means
it's
it's
something
that
that
each
team
will
have
to
stick
to
and
it
may,
when,
whatever
they
did,
they
have
like
a
pressure
on
delivering
their
to
feature
that
their
primary
fee
shirt
and
they
would.
They
would
probably
look
for
things
that
they
they
want
to
put
in
the
back
so
easy.
C
B
Yeah,
that's
a
very
good
point:
I'm
just
trying
to
come
up
with
ideas
of
how
we
can
create
that
feedback.
Loop
right,
so
you
know,
doesn't
necessarily
mean
there's
one
way
to
do
it
or
perhaps
there's
more
than
one
way
to
do
it
and
it
will
lecture
and
he
said
like
there
are
probably
different
types
of
organizations
that
have
different
work
mode
kind
of
let's
say
motivations
to
to
have
that
feedback
loop.
So
you
know
what
other
means
are
there
that
perhaps
you
Joel
or
Adrian
see
in
terms
of
what?
B
What
information
can
you
provide
back
that
can
help
and
whether
that
information
can
be
and
I'm
talking
about
systemic
information?
I'm?
Not
talking
about
you
know
individual
information
because
I
know
Jill
is
passionate.
I
know,
Adrian's,
passionate,
I'm
passionate.
We
all
spend
our
time
and
in
kind
of
driving
value
in
the
ecosystem.
But
what?
What?
What
methods
do
you
see
for
are
feasible
for
your
organization?
B
That
can
you
know,
drive
a
bigger
contribution
if
you
will,
from
your
ecosystem
of
you,
know
thousands
of
Plus
developers
or
my
my
ecosystem
of
a
2,000
plus
developers
to
effectively
have
some
some
feedback
cycle
and
just
to
throw
some
examples
out
there.
Like
do
you
track
your
version,
your
note
version
history
and
your
performance
characteristics
of
your
servers
and
you
historically
look
back
at
those
and
say:
oh,
you
know,
we've
improved
or
not
improved.
Do
you
look
at
the
amount
of
error
rates
you've
been
getting
on.
B
You
know
sure
versions
of
your
code,
but
also
versions
of
what
you're
running
under
your
code,
libraries,
the
tooling
and
so
on.
Is
there
any
way
to
anonymize
that
data
and
perhaps
publish
it
somewhere
or
feed
it
back
to
a
crusted
organization
or
group
part
of
the
node
ecosystem,
so
that
that
can
help
them
do
something
better
about
it
or,
like
we
said,
is
there
wait
for
an
organization
to
take?
You
know
bleeding-edge
and
run
it
and
try
to
do
some
feedback
cycles
around
that
I'm,
just
like
ideation
here
on
the
fly?
B
E
A
lot
of
these
organizations
are
very
regulated,
and
not
this
not
have
to
say
that
you
know
performance
data
is
sensitive
data
at
all,
but
there's
an
investment
on
their
side,
at
least
to
be
able
to
expose
that
data,
and
so
an
incentive
for
them
to
do
that.
I,
don't
know
what
that
would
be
and
how
they
would
do
that
be
able
to
do
that,
but
in
the
reverse
direction.
E
Getting
back
to
your
previous
question
about
you
know,
Joel
said
said
it
best.
You
know
the
magic
word
value
add
if
you
could
show
that
you
know
doing
this
kind
of
systematic
testing
earlier
on
can
essentially
save
you
a
lot
of
money
down
the
line
before
you
get
out
to
production,
where
you
actually
got
full
production
workloads
and
then
you're,
seeing
performance
degradation
out
there
I
that
might
be
a
good
enough
incentive
for
them
to
invest
in
it
and
then
to
get
back
out
I.
B
E
Know
so
maybe
this
is
a
bit
biased,
but
in
a
lot
of
the
consultancies
are
here
working
directly
with
the
leaders
here
who
are
running
programs
trying
to
achieve
some.
You
know
better
performance
unlock
for
productivity,
reduced
costs,
partnering
with
some
of
those
organizations
make
help
with
some
traction
could
be
late
into
yeah
could
be
a
leading
to
exactly,
and
then
that
could
be
the
same
channel
that
you
get
the
information
out
so.
F
That's
I
what
we've
been
talking
about
this
that
actually
leans
into
something
I've
been
thinking
about
urged
like
since
Dan
wrote.
No,
no
just
follow
appoints
what
if
this
was
a
foundation,
a
foundation
provided
service,
no
necessary
foundation
provided,
but
basically
it
was
a
way
for
the
foundation
to
engage
with
enterprises
at
a
higher
level
and
help
connect
them
to
people
who
can
actually
do
this
effectively?
Would
that
make
sense,
or
does
that
sound
attractive
like
would
that
be,
and.
E
I
think
I
think
there's,
there's
gonna
be
someone
at
the
organization's
you're
trying
to
get
to
as
well
as
organization
is
working
with
those
organizations
directly
as
consultants
or
or
in
whatever
capacity
they
are
getting.
Those
people
together,
kind
of
like
we
are
doing
right
now,
would
be
definitely
a
good
way
to
do
this
in
some
sort
of
official
capacity
for
sure
yeah
yeah.
B
Google
recently
published
a
couple
of
case
studies
about
performance
numbers
and
impact
on
cost
and
we've
compared
that
to
our
internal
kind
of
you
know,
customer
acquisition
cost
and
cost
revenue
cost
and
all
those
kind
of
things
and
and
the
way
we
were
looking
at
it
and
specifically
just
to
be
clear
from
like
web
performance
perspective.
In
terms
of
you
know,
every
second
cost
us
X
amount
of
dollars
or
Y
amount
of
dollars,
but
the
same
can
be
said
for
our
infrastructure
and
our
operational
performance
needs
just
to
look
at
it
from
you
know.
B
If
I
upgrade
a
version
of
node
and
I
get
10%
performance
improvements
or
even
5%
performance
improvements,
that's
gonna
mean
millions
of
dollars
to
me,
but
I
have
no
capacity
as
a
business
to
go
and
do
that
on
my
own.
If
I
don't
have
that
driver
eternally
to
force
me
to
do
it
and
your
point
journey
if
if
the
Foundation
provides
at
least
the
inception
point
of
that
or
the
guiding
hand
to
that,
then
you
know
it's
all
about
framing
it
in
the
in
the
business
context,
given
what
the
business
is
about.
B
C
Yes,
all
so
on
for
e-commerce,
for
us
for
it's
less
about
individual
performance,
it's
more
about
like
how
many
calls
do
we
have
to
tear
up
to
to
allocate
to
run
this
app
in
order
to
serve
yeah
so
and
as
far
as
performance
goes,
you
know
we
we
do,
we
do
look
at
them
and
we
do
see
the
try
try
to
see
how
much
improvement
for
every
major
note
version
we
get,
but
to
your
question
we
we
don't
actually
track
them.
So
that
is
a
good
point.
C
We
pass
should
kind
of
like
keep
this
this
history
record,
but
but
so
far
every
major
note
version.
We
see
a
huge
at
least
double-digit
improvement
in
terms
of
performance,
especially
from
version
6
to
version
8,
and
you
know
we
look
at
memory,
usage
and
CPU
usage
and-
and
we
do
it
so
far,
I
think
the
for
us
is
that
you
know
we
look
at
me.
The
more
important
thing
for
us
is
a
hub.
Is
there
any
errors
like
you
mentioned,
and
we
do
track?
C
B
A
Feedback
point:
it's
a
good
thing
for
us
to
get
into
care
back,
as
you
know,
recommendation
from
this
group
that
that
in
action
takes
place
so
I'd
be
remiss
if
I
didn't,
you
know,
poke
everyone
a
little
bit
with
the
stick.
You
know
Adrienne,
Joel
and
Achmed.
All
three
of
you
are
not
members
of
the
no
dis
foundation,
and
so
one
of
the
best
ways
to
ensure
that
services
are
delivered
on
behalf
of
the
end
user
community
is
to
also
encourage
that
your
organization's
do
join
the
foundation.
A
There
are
some,
you
know
relatively
easy,
onboarding
points.
You
know
you
don't
have
to
go
in.
You
know
all
into
you
know
the
top
tier
commitment
to
to
be
a
valuable.
You
know
contributor
there
I.
Could
you
know
to
bootstrap
what
you're
proposing
you
know
rather
than
you
know,
setting
are
sort
of
way
marker.
As
you
know,
if
the
foundation
has
this
thing,
you
know
them
will
start
it.
What
what
we
might
want
to
do?
You
know
with
this.
You
know
enterprise
focus
group.
A
Is
you
know
in
these
sessions
set
ourselves
a
deadline
of
we're
gonna
collect
some
data
together
right
now,
maybe
for
the
next
session.
You
know
we
could,
you
know,
do
a
little
homework
and
you
know
collect
some
data.
You
know
put
some
stuff
in
the
pipeline.
Get
some
get
some.
You
know
actual
sort
of
under
under
usage.
A
You
know
trials
of
what
the
data
looks
like
of
how
labor-intensive
it
is
or
isn't
to
put
in
to
our
CI
CD
pipelines.
And
you
know,
then
you
know
it's
it's
that
you
know
initial
concrete
version
of
the
thing
that
you
know
it
is
often
the
spark
that
you
know
really
gets
the
ball
rolling.
So
you
know,
rather
than
you're
talking
about
the
potential
future.
You
know
I
think
we
have
an
opportunity
with
this
group
to
you
know
really
begin
some
action
that
we
can.
A
You
know
snowball
upon
and
you
know
make
make
recommendations
you
know.
For
example,
you
know
this
groups,
the
user
feedback
group
just
helped
land
with
the
cooling
focus
team,
make
turkey
and
some
other.
You
know
user
user
convenience
functions
that,
were
you
know,
constantly
being
leveraged
from
NPM
and
those
are
landing
NFS,
so
folks
are
listening
and
I
think
this
is
a
fantastic
way
to
have
a
big
impact.
C
Yes,
thank
you.
Dance
toast
was
a
very
clear
point
in
on
a
foundation
thing.
Yeah
I
should
join
personally
myself
and
have
I've
been
talking
to
our
leaders
on
having
a
Walmart
joining
the
foundation,
but
I
always
comes
back
to
that
question.
All
about
you
add
right
and
I
have
another
meeting
attend.
I
have
to
go,
but
I
want
to
say
something
really
quick
about
you
know
at
which
the
training
thing
so
in
in
the
first
year.
One
of
the
biggest
thing
that
I
struggle
with
is
to
make
our
developer
understand.
C
Uncaught
exception
is
like
crashing
Java.
You
know
I'll
develop
which,
which
one
I
deploy
apps
to
production,
with,
with
the
console
like
constantly
spinning
out
uncaught
exception
like
one
per
second,
and
like
guys,
you
you
have
to
understand
this
note
may
not
crash
when
there's
an
exception,
but
it's
kinda,
like
you
know
that
no
point
extorting
in
Java
you're
AB
crash.
You
have
to
do
something
about
these
errors,
so
that's
one
of
the
biggest
struggles
I
always
had
in
the
beginning.
Even
now,
sometimes
yeah
so
I
have
the
trouble.
Sorry
guys,
yeah
yeah.
Thank
you.
B
B
Think
if
we
follow
up
with
some
potential
homework
of
what
data
can
we
collect
and
what
information
can
we
come
back,
but
it
doesn't
have
to
be
systematic
today,
but
you
know
I
for
myself,
I'm
gonna
take
away
just
coming
back
with
some
insights
about
the
penetration
of
nodejs,
an
organization.
The
number
of
what
versions
are
we
running
and
and
how
long
can
they
be
running.
A
B
Don't
know
if
I
can
get
performance
numbers
historically,
but
maybe
something
can
look
at
it
and
specifically
like
we
actually
I
was
gonna
mention
this,
but
we're
running
low
on
time.
We
built
something
internally
where
we
go
through
every
single
github
repo
in
our
org
and
look
at
every
single
dependency
that
we
have
in
our
organization.
You
know
we
know
what
we're
actually
you're
lying
on
the
mat,
the
most
to
run
our
business.
You
know-
and
you
know,
spoilers
various
parts.
B
Yes
as
a
off
the
topic
example,
but
there
are
things
I,
don't
know,
I,
don't
even
know
right.
So
having
created
that
kind
of
library,
scanner
just
to
understand
what's
out
there
and
what
developers
are
doing
has
actually
opened
up
our
eyes,
a
lot
of
things
and,
most
importantly,
to
the
versions
that
people
are
using
because
not
everybody's
on
the
latest
version,
which
means
security,
vulnerabilities
and
other
issues
do
come
up.
B
But
that's
something:
I
was
gonna
plan
to
open-source
and
trying
to
make
it
anonymous
in
a
way
so
that
you
can
run
it
in
your
own
organization
and
then
for
yourself
have
that
visibility.
But
then
maybe
that's
something
we
can
publish
and
share
back
within
this
group
and
say
look.
This
is
what
our
business
is
run
on.
This
is
hundreds.
A
So
the
one
other
thing
that,
before
you
close
this
out
up
and
one
of
the
things
that
you
know
we
could
potentially
from
the
community
committee
no
Jess
bring
back
to
this.
You
know
enterprise,
user
feedback
focus
group.
You
know,
there's
been
a
lot
of
effort
put
into
revising
and
updating
the
documentation,
that's
still
a
work
in
progress,
but
one
of
the
you
know
potential.
Since
you
know
the
the
docs-
and
you
know
the
the
context
of
getting
you
know,
sort
of
the
enterprise
team
you
know
off
the
ground
is
so
important
in
this
context.
A
You
know
one
of
the
things
that
that
you
know
could
be
really
valuable
is
providing
some
of
that
early
feedback.
So
we
could
invite
a
documentation
and
and
website
redesign
team
member
to
this
session
to
share
some
of
the
work.
That's
that's
going
on
there
and
provide
feedback
to
you
know
how
this
is.
This
is
or
isn't
valuable
to
you
know.
Enterprise
needs
I'm.
F
Sitting
right
here,
Dan,
actually
a
lot
of
what
you're
saying
is
very
spot-on,
with
our
plans
for
the
website
redesign
work,
we're
expanding
out
to
guides
beyond
Docs
and
have
a
kind
of
an
RFC,
I
guess
out
for
people
to
come
and
help
us
build
out
a
bunch
of
the
guides.
We're
looking
for
two
or
three
people
to
begin
that
process,
just
to
kind
of
get
a
few
solid
guides
out
of
the
way
and
kind
of
build
a
pound,
a
path
down
for
for
people
to
to
follow.
F
B
Also
add
one
more
homework
topic
for
for
our
next
session.
Whatever
that
would
be
is
to
reach
out
to
your
peers,
who
are
also
in
big
organizations
and
are
depending
on
no
js'
and
let
them
know
that
we
are
having
this
dialogue,
because
they
might
also
be
interested
in
some
form
of
contribution
and
or
helping
the
feedback
cycle
of
we
mentioned.
I.
Think.
D
In
order
to
do
that
so
Ahmed
first
good
job
in
reading
this
meeting
so
first
try
to
open
an
issue
in
in
a
user
feedback
so
that
that
will
be
great
because
from
there
we
can
take
it
and
we
can
show
to
the
other
people
and
at
least
like,
for
example,
for
Intel
and
the
topics
are
great,
but
the
people
they
need
to
know.
Usually
they
are
looking
on
YouTube
or
and
sometimes
on
the
MD
files
on
the
notes
and
that's
it
yeah.