►
From YouTube: Package Maintenance Team meeting - 14 July 2020
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
A
A
B
Sorry
I'm
just
kidding
kind
of
yeah
yeah,
that's
pretty
much
just
looking
for
a
feedback.
I
think
there
might
be
enough
approvals.
I
haven't
checked
in
a
little
bit,
but
yeah
it
was
just
yeah
just
looking
for
any
thoughts
or
otherwise.
I
guess
would
be
good
to
use
it
to
start
helping
out
some
of
the
teams
I've
been
asking
us
for
help
like
Axio
sand,
node
fashion
groups
like
that
projects
like
that,
okay.
A
C
B
Mm-Hmm
I
should
have
a
question
on
that
approval.
Is
that
a
is
that
for
someone
to
land
their
own
PR?
Is
that
or
any
PR
cuz
I
did
merge
one
yesterday
that
wasn't
seven
days
but
had
four
approvals.
I
was
just
fixing
a
link
in
the
readme,
so
hopefully
that
wasn't
overstepping
stepping
or
overstepping
my
bounds.
I.
A
B
That's
that's
what
it
was
that
might
have
been
an
okay,
but
it's
still
good
to
clarify
in
general,
so
yeah
I'll
take
another
look
at
the
PR
and
you
know
just
make
sure
I
didn't
leave
any
more
of
my
own
typos
in
there,
but
yeah.
That's
good!
Then
I'll,
merge
and
also
start
sharing
that
information
with
some.
The
related
issues
of
teams
asking
us
for
something.
B
A
E
A
B
A
F
Quick
update-
and
this
is
there's
been
some
progress
made
its
couple
POS
been
motion
to
would
be
there's
still
more
progress
to
be
made.
I
think
I
need
to
go
fix
up
on
one
of
the
peels.
There's
PRI
bring
dependents
as
well.
Tom.
Are
you
okay?
If
I
merge
that
in
a
local
enough
appeal
for
your
twinks,
the
one
over
independence?
That's
mentioning
I'm
an
MP
mo
e
I.
Think
you
yeah.
You
remember.
E
E
F
A
F
A
A
B
A
No
okay:
let's
move
on
to
the
next
one,
which
is
next
steps
on
support
levels
and
package,
Jason
that
one
I've
got
I.
Think
the
the
first
version
of
the
support
to
landed
I
do
need
to
look
back
in
that
it
doesn't.
It
has
a
dependency
which
doesn't
support
eight,
so
the
CI
tests
are
failing,
but
I
think
we
agreed
that
was
okay,
I,
just
I'll.
A
That's
where
it
is
it's
under
pkgs
support
to
West
had
started
it
I
just
thought
I
just
added.
We
decide
we
gotten
together
and
agreed
on
what
sort
of
the
minimum
MVP
was
and
that's
its
I
thought.
Still.
A
lot
of
work
could
be
done
so,
like
you
know,
I'm
glad
and
that's
probably,
an
area
we
could
get
somebody
involved
in,
but
it
does
the
basic
validation
and
listing
to
start,
and
then
you
know
we
can
build.
A
A
Okay,
any
other
questions
on
that
one.
A
G
Sir
as
amuse
I
I
opened
that,
because
Dominick
has
made
a
comment
and
actually
run
very
true
like
let's,
we
can't
just
offer
advice.
We
have
to
offer
concrete
examples.
This
is
the
way
you
can
do
it
yeah,
because
that's
often
the
problem
for
a
lot
of
people
like
I,
don't
know
how
to
do
this
thing.
Can
you
give
me
a
good
example?
So
I've
gone
away
of
good
I've,
looked
at
gate
labs
and
don't
forget
hub
actions
at
a
moment.
G
What
perhaps
we
could
do
is
take
one
of
the
pkg
just
just
like
dependence
yeah
and
see
I,
CD
it
to
death
and
say
so.
We
have
a
good
example:
yeah
we've
github
action
say
this
is
how
we
did
one.
This
is
a
good
example
of
how
to
do
it.
Then
we
have
a
concrete
example,
though
we
actually
maintain,
rather
than
pointing
to
references
for
other
people's
things.
We
wrote
a
CIC
document.
G
I
think
Dominic
is
right.
We
have,
it
has
to
actually
be
opinionated
lots
of
times.
We
don't
like
to
be
opinionated,
but
I
think
it
in
this
particular
case.
We
need
good
examples.
Yeah
like
we
want
to,
like
you
know,
just
you
know,
use
resource
stuff.
We
would
do
on
Travis,
you
know
I've
run
tests,
but
no
we
would
then
we
want
to
spin
it
up
on
docker.
We
want
to
run
it.
We
want
to
see.
Did
some
dependence
found
yeah?
We
wanted
to
see
here,
endpoints
and
so
people
could
take
that.
G
As
you
know,
a
you
know,
a
good
case
example:
I
must
apologize.
Most
of
you
know
to
stuff
I
do
during
the
day.
Time
is
the
traditional.
You
know,
Jenkins
based
stuff
and
now
I'm
looking
into
the
the
more
highly
used
ones.
So
we
we
do
stuff
already
that
everyone
knows
about
like
the
Travers
and
stuff,
but
I'm
really
looking
at
github
actions
because
of
its
docker
integration,
and
you
know
the
fact
that
we
can
run
things.
You
know
it
does
windows
it
does.
Mac
does
lionex,
it
seems
to
be
offering
quite
a
lot.
G
A
G
So
that's
why
I'm
kind
of
suggestion
so
Andrew
if
I
come
along
and
start
raising
PRS
to
CIC
doing
stuff
with
like
currently
I'm
thinking
of
github
actions,
I
mean
they
are
for
open
source
onesie.
You
know
it
is
very
clear
that
it's
free
and
it
does
appear
to
be
quite
performant
I-
think
that
would
be
a
good,
a
good
starting
point
to
go
forwards,
dominate
because
your
opinions,
because
I
when
you
said
they
resonated
with
me
like
hey,
we
need
examples
because
we
will
learn.
That's
the
way
we
learn
like
we.
G
E
Yeah
I
was
also
kind
of
referring
that
to
that
the
very
basic
concepts
we
don't
need
to
go
into
much
detail
in
the
docs,
primarily
because
there's
a
lot
of
authoritative
sources
that
have
been
writing
about.
You
know
what
is
CI,
what
is
C
what's
an
integration
test,
so
we
maybe
don't
need
to
go
into
as
much
detail
in
there.
E
As
for
the
examples,
yeah
I
kind
of
agree,
I
think
it's
a
bit
trickier
to
to
give
CD
examples
for
note
packages
right
because
CDE
of
a
note
package,
the
lifecycle
of
that
ends
with
an
NPM,
publish
and
you're
done
right,
because
you
can't
freely
you
you're,
not
really,
unless
you're
building
something
like
I,
don't
know
ghosts
where
you
have
a
blogging
platform
that
you
deploy
into
a
cloud.
And
you
know
it
combines
all
the
modules
so
yeah,
but.
G
I
think
there
is
the
main
and
I
we
sometimes
do
I
sauce
out,
and
we
say
we
want
this.
You
know
we
want
to
like
talking
in
a
general
border
terms
about
you
know
different
ecosystems.
We
are
the
biggest
ecosystem
and
we-
and
let's
shout
about
something
we
are
the
you
know,
but
this
is
the
biggest
community
in
the
draw-
and
this
is
the
most
common
case
is
to
publish
after
you've
built
to
MPN.
E
G
B
B
Does
it
help
or
is
it
worth
trying
to
understand?
Maybe
the
I
guess
the
audience
were
trying
I
mean.
Obviously
it's
package
authors
but
beginners
advanced
I
mean
I,
guess
anybody
that
has
already
set
it
out,
you
know
would
be
able
to
skim
through
a
document
like
this,
but
someone
who's
never
set
it
up
potentially,
but
you
know
you
could
benefit
from
concrete
examples
as
Glen
is
referring
ya
know.
Is
there?
B
G
I
think
if
we
can
get
the
basics
right,
that,
like
99%
of
people
who
are
just
publishing
everyday
regulus,
things
could
use
and
they
could
use
these
tools
that
are
now
available.
That
would
be
a
big
boost
and
we
can
provide
the
wreck.
Dacians
I
mean
it's
a
big
thing
to
be
able
to
like
compile
your
code.
G
You
know
run
the
tests,
you
do
the
lint
in
and
then
you
know,
do
a
small
deployment
in
in
a
container
and
actually
try
it
out
like
integration
tests,
yeah
and
then
define
these
is
deployment,
but
the
first
part
is
something
to
actually
integrating.
You
know.
Code
is
something
most
people
don't
don't
know
how
to
do
yet
or
is
not
widely.
It
is
not
good
advice.
So
that's
kind
of
the
first
steps
so
yeah.
B
Because
I
guess
to
me
I
think
the
the
if
there
is
kind
of
one
thing
that
could
really
help
I
think
what
most
is
yes,
I
think
is,
is
in
the
PR
process.
Right,
you
know
it's
not
if
we're
kind
of
assuming
github
as
that
common
denominator
and
github
actions,
then
you
know
what
makes
you
feel
more
confident
with
your
own
code
and
other
people's
code
than
seeing
that
nice
green
circle
at
the
bottom
of
your
pull
request
right
yeah,
that
is
you
know,
so
maybe
that's
kind
of
yes
exactly.
G
B
And
then
you
know
we
can
always
add
more
stuff
after
but
yeah
I
think
I
think
having
at
least
something
that
gives
everybody
the
ability
to
at
least
cross
that
you
know
line
for
that
I'll
stone
like
hey.
You
can,
because,
even
when
my
conversations,
unlike
get
everybody,
always
says
they
create
their
own
bridges,
even
a
lot
of
people
that
work
in
their
own
projects.
B
B
G
G
Through
the
docks
is
making
sense,
it
looks
like
olds
are
a
lot
of
proper
promises.
It
seems
to
be
prominently
free
for
open
source,
which
is
like
a
big
thing.
The
others
are
not
saying
that
that
that's
a
big
deal
and
that's
why
the
advice
you
know
Travis
is
free
for
open
source,
but
it's
not
really
a
CI
CD
platform,
whereas
they're
holding
out
that
promise
for
github
actions.
It
is
now
whether
you
use
their
runners
or
you
decide
to
use
your
own
runners.
This
is
what
the
end
goal
is
and
I
think
your.
C
B
B
A
G
Yes
and
those
are
the
advanced
features
we
need
to
start
working
on
for
the
more
challenging
things
and
I'm
gonna
like
hold
my
hand
up,
say
that
I
worked
in
a
dotnet
shop
by
eight
nine
years
ago
and
I
can't
even
remember
I,
just
I.
Just
can't
remember
what
to
do.
You
know
like
so
yeah
that
we
I
acknowledge
that,
and
there
may
be,
as
you
say,
more
advice
than
you
know.
We
may
need
to
once.
We
get
the
basics
here.
G
B
B
G
F
B
To
the
team,
okay,
we're
doing
it:
okay,
yeah
yeah.
G
A
E
E
C
E
E
Cozy
top
right,
so
you
would
start
off
with
two
strategies,
essentially
the
way
all
versions
or
LPS
versions.
The
all
versions
would
basically,
if
you
include
one
of
the
following
files
right
so
greater
or
equal
ten,
it
will
mean
that
it
will
import
all
the
node
versions
since
version
10
and
you
will
be
testing
in
those,
whereas
the
strict
would
import
just
the
versions
that
are
going
to
be
LTS,
and
it
would
also
include
the
current
version.
Even
if
that
is
not
a
yes.
E
So
a
couple
of
months
ago,
in
GTE,
twelve
would
have
brought
you
12,
13
and
14.
For
a
couple
of
weeks,
right
13
is
allowed
to
drop
off
the
obvious
list,
whereas
in
the
case
of
all
whatever
you
import
there,
that
list
will
only
grow.
It
will
never
get
removed
any
versions
right.
So
so
they
commitment
to
its
the
it
depends
on
your
support
strategy.
E
Right,
if
you,
if
you
support
all
know
your
support,
Homo
Persians,
whereas
if
you
support
a
yes,
then
we
only
include
current
for
the
sake
of
testing
and
also
added
a
strict
folder
under
LTS,
and
the
I
hid
it
inside
of
the
S
rather
than
at
the
top
level.
Saying
something
like
out
the
air
district,
primarily
because
the
recommended
approach,
I
would
say
should
be
that
you
should
be
testing
in
LTS
plus
current,
even
though
your
support
policy
says
that
you
only
support
LTS
right.
E
So,
even
though
we
only
support
LTS
from
the
perspective
of
the
end
users,
you
still
want
to
be
running
tests
in
the
current
version.
Yeah,
because
that
you
know
gives
you
feedback
gives
you
feedback,
whereas
strict
only
includes
the
versions
that
are
actually
LTS
or
potentially
have
expired
right.
So
this
does
not
include
14
yet,
for
example,
because
14
will
only
become
LTS
in
so
yeah.
That's
that's!
Roughly
the
story.
A
E
So,
but
that's
the
thing
between
what
you're
actually
testing
in
and
what
you're
supporting
it,
and
if
you're
saying
that
I'm
supporting
the
LTS
and
today
the
actual
LTS
versions
are
tenant
well
yeah,
so
you
wanna
test
in
10
and
12
and
14,
because
14
is
upcoming
here
here
and
by
the
time.
10.
Support
for
10
ends,
because
at
the
moment
to
change
that
to
stop
testing
in
10.
E
You
do
need
to
bump
the
major
version
right,
because
if
you
drop
support
for
a
specific
node
version,
then
your
needs
to
bump
your
major
version,
which
means
that
you
have
to
make
an
explicit
opt-in
to
drop
the
gte
10
right.
So
if
today,
if
you
were
starting
today,
you
would
go
with
gte
10
and
this
would
always
include
10
12,
14
16,
but
by
the
time
10
expires,
you're
still
supporting
gte
10
and
you
bump
a
major
version
in
which
case
you're,
also
okay,
to
change
to
GT,
12
or
whatever.
Okay.
A
E
E
But
that
means
that
for
people
who
use
that
kind
of
a
strategy,
it
means
that
as
soon
as
10
goes
out
of
support,
they
drop
10
and
they
do
not
release
or
do
not
have
to
release
major
digit
of
testing
in
it
and
there's
a
sort
of
and
I
can
see
people
going
against
that
kind
of
a
recommendation.
If
we
want
to
it's,
it's
yeah,
it
probably
makes
sense,
but
it
is
something
to
discuss
again
whether
we
want
to
provide
such
a
configuration,
which
automatically
drops
expired
versions.
I.
A
C
B
E
E
I
think
it
says,
force
but
I
think
it's
a
fair
suggestion
and
as
a
package
maintainer
I
would
I
could
probably
even
subscribe
to
that,
and
we
could
say
that
he
hears
our
policy.
We
automatically
drop,
expired
LTS,
no
aversion
support
without
releasing
major
and
I'm
sure.
There's
people
who
would
do
and
would
want
to
do
that,
but
there's
being
a
lot
of
contention
around
that
around
expiring
support
for
specific
node
versions,
even
though
they're
no
longer
supported
without
bumping
major.
D
E
E
A
We
might
want
to
talk
to
keep
talk
about
that
more
I
mean
I.
Don't
think
people
are
like
I
just
keep
wondering
if
we
can
get
there,
but
I.
Don't
think
people
are
in
agreement
enough
to
add
it,
yet
that
the
flip
side
that
I've
seen
which
we
saw
in
the
napi
one
is
like
we
don't
actually
break
anything
right
like
so.
We
did
a
version
three
just
because
we
want
to
drop
support
for
eight.
A
Then
somebody
said
oh
well,
there's
this
thing
that
wasn't
fixed
and
you
know
I'm
still
using
eight,
but
it's
not
fixed
into
and
we're
like
well,
but
it's
it's
fixed
in
three
and
three
doesn't
break
you,
so
you
could
use
three,
but
that's
the
one
that
officially
dropped
support
for
eight
and
like
so.
If
we
hadn't
been
forced
to
make
a
new
version
to
drop
support
freight,
then
they
wouldn't
have
been
in
that
you
know
it
would
have
been
like
hey.
A
We
know,
like
they've,
made
the
choice
they're
using
something
which
is
unsupported,
because
we've
said
we
don't
support
eight
anymore,
like
it
doesn't
matter.
If
it's
like
two
said,
it's
supported,
eight
well
yeah.
It
did,
but
we
no
longer
do
right,
like
we
only
support
and
we're
only
actively
maintaining
the
ones
that
match
the
card
LTS,
but
having
a
dad
to
do
that
bump
actually
made
it
worse
and
we
actually
did
a
release
on
two
for
something
which
was
unsupported
and
it's
like.
So
there
are
cases
where
that's
not
helpful
right,
but.
E
I
think
I
think
that
even
more
so
this
speaks
to
the
idea
that
if
you
import
your
tribes
config
to
be
GTE
eight
and
then
you
drop
support
for
eight,
even
though
you
don't
support
it,
you
can
still
keep
running
tests
and
when
you
actually
achieve
a
point
where
you
break
eight,
you
will
know,
because
you
have
eight
tests
for
eight,
and
that
is
when
you
remove
it
from
Travis
to
progress,
and
that
is
when
you
tell
people.
Okay,
we've
told
you,
we
don't
support
it.
A
E
No
yeah,
I
think
I
think
it's
worth
providing
the
config,
which
automatically
drops
the
unsupported
versions
and
I
guess
it
would
go
under
LTS
folder
and
it
would
have
something
like
or
current,
documenting
or
active
that
llamo
or
whatever,
with
the
keyword
that
we
chose
right,
maintained
our
channel
and
then
it
would
only
give
you
the
currently
maintained
healthiest
versions.
So
in
LPS
it
would
give
you
health
es
plus
current
in
LTS
strict.
It
would
only
give
you
the
current
strict
in
all
it
would
in
all.
It
does
not
make
sense,
I
guess,
because
no.
E
E
I'm
in
this
situation,
for
example,
I,
my
favorite
preferred
testing
framework
is
happy,
lab
and
officially
does
not
support
notes
then,
but
I
think
it
still
runs.
Its
tests
intend
and
I
use
it
in
10
and
it
does.
It
didn't,
do
anything
specifically
to
break
them
so
right,
yeah,
I,
I
I'd.
Rather
do
it
in
a
separate
issue
and
a
separate
they
are
yeah.
B
Not
not
to
add
more,
oh
it
just
for
a
thought,
I
wonder:
there's
a
distinction
between,
in
terms
of
communication,
documentation,
support
between
dropping
aversion
versus
like
formally
adopting
a
version.
You
know
where
yeah
in
the
spirit
of
bike
shedding
yeah,
add
more
permutations
today,
the
matrix
of
things
to
discuss,
but
anyway,
yeah
just
just
cuz
I,
know,
there's
always
lots
of
good
conversations
around.
B
Like
you
know,
dear
dependencies
or
even
your
runtime,
because
this
happens
a
lot
like
typescript,
especially
you
know,
do
those
influence
the
versioning
of
your
own
package
so
and
I
think
there's
a
lot
of
nuance
there.
So
you
know
maybe
there's
not
a
one-size-fits-all,
but
there's
you
know,
maybe
the
in
the
spirit
of
this
group.
Maybe
the
outcome
is
like
well,
here's
the
you
know.
Maybe
there's
like
a
good,
like
you
said:
maybe
there's
an
option
where
you
do
always
bump
it.
B
A
A
A
Because
I'm
anyway,
let's
take
that
to
a
separate
issue,
be
interesting
to
work
through
the
cases
where,
like
you
know,
I
it's
you
know.
What's
the
downside,
if
you
haven't
bumped
your
major
and
what's
the
downside,
if
you
have-
and
you
know
just
have
that
discussion
figure
out
whether
we
think
having
that
as
an
option
that
you
can
do
automatically
helps
or
hurts
things.
A
Okay,
so
sorry
back
to
that
one,
any
more
comments
on
that
I
think
it's
anything
else
you
wanted
to
highlight,
like
I,
think
that
I
now
understand
it.
I
was
kind
of
wondering
a
bit
because
I
was
expecting
be
more
of
the
just
include
one
file,
and
now
you
don't
actually
have
to
do
anything,
but
we
discussed
that
so
I
mean
in
the
context
of
you
need
to
make
a
change.
What
you've
got
makes
sense
to
me.
A
B
I'm
not
sure
if
I
missed
it
at
the
beginning
or
but
was
there
any
feedback
or
yes
or
anything
to
share
coming
out
of
collab
summit
and
the
conference
a
couple
weeks
ago
from
this
group,
I
wasn't
able
to
attend
so
I
wasn't
sure
how
what
but
I
know
we
presented
and
yeah
I
was
curious.
How
that
turned
out.
B
E
A
G
It
went
well
I,
wouldn't
say:
there's
anything
of
extra
special
interest,
I
think
when
we
got
together
afterwards.
We
talked
a
lot
of
things.
You
know.
Well,
there
were
very
common
themes,
one
of
one
of
them
really
is
we
have
to
get
that
Enterprise
Engagement,
going
because
when
we're
holding
these
talks
were
really
talking
to
people
that
already
agree
with
us
and
we
need,
as
michaelis
said,
to
get
out
and
do
a
bit
of
vandalizing
I'm,
certainly
doing
it
in
my
organization.
Now
that
I
think
and
that
I'm
Darcy's
very
accurate
mark
of
litter.
G
G
A
One
of
our
goals
should
be
so
you
want
to
add
the
Kate,
for
you
know
your
organization
supporting
XY
and
Z,
like
efforts
of
the
package
maintenance
working
group,
the
Open
Jazz
foundation,
the
what
can
we
put
together
based
on
you
know,
you've
obviously
got
some
experience
working
at
your
company
trying
to
do
that
now,
so
maybe
trying
to
capture
and
think
about
will
work.
What
didn't
work
I
know.
C
A
G
Believe
so,
yes
and
I
think
that
some
of
the
things
I
want
to
cover
later.
You
know
like
if
I
bring
people
in
to
talk
into
my
institution
and
I'll,
bring
you
in
at
a
high
level.
There
be
just
a
leaders
meeting
and
say
you
know:
I
work
at
this
company.
We
have
these
projects,
you
can
come
and
work
on
it
and
with
most
enterprises,
large
companies
they
like
working
with
other
recognized
large
companies,
so
cerebral
dhalsim
does
I
am
from
NPM.
These
are
my
projects
that
I
require
help
on.
G
You
are
free
to
come
along
that
sort
of
thing,
and
it
might
be
an
idea
that
we
actually
set
goals
for
people
like
you
need
to
get
ten
people
in
the
open
source
or
20
people
from
your
organization.
I,
don't
know
what
I
what
we
I
have
family
experience
is
actually
setting
targets
is
helpful
because
then
you
know,
if
you
worked,
we
didn't
work.
You
know,
I
have
hard
targets.
We
want
to
get
this
many
people
involved.
We
want
to
see
some
engagement,
yeah.
A
A
G
Other
thing
is
I.
Think
mother
things
we
do
see
from
the
express
triage
group
is
not
not
everyone
that
joins
is
like
the
active
member
come
comment.
You
often
make
like
you
may
get
members,
but
how
many
of
them
become
truly
active
long-term?
That
figure
that
you
know
they
stay
and
they
contribute
long-term.
B
G
Circle,
it's
both
a
case
of
time
and
permission
so
from
organization.
It
was
a
case
of
they
had
to
go
around
that
did
I,
create
a
process
and
there's
a
lot
of
companies.
They
involves
lawyers
like
what
are
they
up
to
terrified?
You
know
like
what
is
this
open-source
thing?
What
do
you
mean
we're
already
using
it.
A
And
I
think
they're
like
one
of
the
talks.
There
is
the
interesting
it
was
basically
making
the
case
that
we
don't
necessarily
want
companies
to
just
provide
money.
The
amount
of
money
we
invest
in
open-source
is
actually
pretty
small
compared
to
the
amount
of
spent
internally,
but
that
really
it's
beneficial
and
also
from
the
business
perspective
you
get
more
than
just
like.
If
you
just
provide
money,
you
don't
you
maybe
support
you're
using,
but
you
don't
get
all
the
benefits
you
can.
A
B
B
B
I
guess
sure,
to
turn
that
money
to
like
a
pool
that
people
can
draw
out
of
and
contribute
to,
it
still
is
really
honor.
If
you
don't
have
the
eyes
on
the
project
in
the
first
place
to
incentivize,
you
know,
if
you
don't
have
anybody
looking
at
a
project
with
you
know,
say
some
money
to
actually
give
away
the
money
you're
just
gonna
sit
there
yeah.
A
It
depends
there'll,
be
some
don't
you
know,
there's
certainly
be
some
groups
or
some
some
things
that
you
know
you
kind
of
want
to.
You
know
turn
turn
it
into
a
full-time
job.
In
other
cases,
people
don't
want
it
as
their
full-time
job,
and
so
they
may
be
more
of
a
community
than
money
to
help
them
do
it
full-time,
but.