►
From YouTube: Package Maintenance Team meeting - Feb 25 2020
Description
A
A
B
B
A
lot
of
our
discussions
were
around:
how
can
a
module
author
test,
whether
their
changes
are
going
to
impact
modules
that
depend
on
us,
and
we
had
a
lot
of
discussion
on
how
we
could
approach
it.
There
was
a
suggestion
of
building
a
prototype,
the
first
step
of
that
being
that
you
just
created,
maybe
a
small,
CLI
or
very
small
tool
that
would
just
get
the
list
of
the
modules
that
depend
on
a
given
module
and
then
we
would
work
from
this
and
then
we
had.
B
B
C
D
B
F
Had
added
back
to
the
agenda?
Yes,
okay!
Yes,
this
was
about
getting
the
Travis
official
Travis
includes
and
that
we
can
share.
If
so
that
people
can
use
those
and
automatically
opt
in
to
the
updates
according
to
their
policy
and
possibly
extend
that
for
other
CI
systems
like
get
up
actions
or
circles.
Yeah.
F
B
F
F
A
D
F
A
F
F
A
F
F
A
F
The
fact
that
difference
guys
would
probably
be
moving
at
different
speeds
and
would
have
different
different
layouts
and
different
expectations
of
how
the
includes
work
so
I
don't
know,
I,
don't
know
how
would
this
be
achieved
in
in
in
github
actions,
but
in
some
cases
they
might
be
expecting
files
at
the
root
of
the
rip
or
something
like
that
right.
So,
basically,
if
you
have
a
repo
Persei,
you're
free
to
do
any
layout
for
each
CI
independently,
without
enforcing
any
kind
of
structure
and
without
affecting
other
configurations
for
other
CIS
right.
A
F
Yeah
and-
and
this
is
a
very
long-term
decision
right,
because
if
you
tell
somebody
take
this
Travis
include
and
you
will
always
get
the
latest
version
of
node
automatically,
then
you
know
it
would
be
tricky
to
to
change
that
and
let
everybody
make
everybody
upgrade
or
change
and
point
point
to
a
new
name.
If
we
decide
that
it's
no
longer
good
enough,
so
a
separate
repo
Persei
is
probably
a
bit
a
bit
more
flexible
right.
F
So
off
the
top
of
my
head,
three
right
travis
circle
and
get
up
actions,
but
I
have
no
idea
what
other
ones
we
might
have,
but
also
this
does
bring
back
to
the
namespacing
discussion
with
the
PK
gjs
and
maybe
Noda
has
grown
out
a
single
organization
right
like
if
we
had
an
organization
called
no
JSC,
I
setup
tools,
oh
gotta
knees,
saying
Glenn
needs
to
be
promoted,
don't
sure
right.
So
if
we
had
an
organization
called
node
CI
tools
for
packagers,
that's
for
maintainers
give
it
a
long
name
description'.
F
You
know
if
we
had
a
separate
organization
where
you
could
just
put
in
a
repository
for
CI
two,
and
it
will
be
an
organization
dedicated
to
CI
configs.
Maybe
you
know
just
namespacing
that
would
be
easier
and
it
also
sort
of
fixes
also
the
team
structure,
permission
structure
and
management.
So
it
does
add
complication,
but
it
does
also
makes
organizing
things
easier,
but
also
at
some
point
note
may
be
probably
will
grow
out
a
single
organization.
So.
A
Yeah
at
this
point,
I
think
you
know
looking
at
the
discussion
on
the
other
issue:
I'm,
not
sure
lots
and
lots
of
orgs
well
like
asking
the
the
the
moderation
team
to
look
after
multiple
them
yeah
it
just
it
just
does
add
like
we
discuss.
We
do
just
need
to
keep
in
mind.
It
does
add
overhead
for
every
other,
org
everywhere.
F
F
F
F
A
G
A
You
know
we
can
probably
pick
a
name
that
would
be
something
like
CI.
Configs
could
be
like
a
repo
name
if
it
was
common,
yeah,
okay,
okay,
so
we
will
say:
let's
get
the
lists,
options
and
the
issue
and
then
get
input
from
the
broader
team.
And
then
we
can
see
Explorer,
the
feasibility
of
which
everyone
we
think
is
preferred.
C
A
A
A
Okay,
the
next
one
is
document
current
top
ten
express
issues.
This
is
quite
early
for
West,
so
not
surprised
to
you
can't
make
it
that's
why
we
have
different
times.
I,
don't
think,
there's
anything
to
new
on
this
one
I
don't
know
if
anybody
has
any
updates
on
the
Express
triage
front
or
I
guess
Bethany.
Sorry,
Glen
go
ahead
so.
D
Yes,
Archie,
the
triaging
is
actually
working
quite
well.
There's
a
couple
of
guys
John
charged
myself
with
shipping,
and
we
actually
there
aren't
that
many
issues
raised
on
Express.
To
be
honest,
they
usually
like
what
am
I
doing
wrong
and
we
find
an
example
here.
You
go,
you
do
it
this
way
and
people
are
pretty
good.
Actually,
the
number
of
open
issues
is
declining
and
that
seems
to
be
working.
D
The
triage
in
itself
is
working
both
ways
and
I
are
just
going
through
a
patch
of
extra
responsibilities
that
are
distracting
de
triumphe
vocations,
which
is
slowed
us
down
immensely
we've
been,
we
have
been
chatting
or
flying
the
both
of
us
in
the
same
situation,
which
is
free,
just
temporary
and
we'll
be
getting
back
to
where
Z
is
top
ten
issue,
which
basically
the
big
issue
at
that
moment.
Is
it
just
takes
so
long?
The
crew,
through
all
the
issues
on
Express,
but
actually
the
fundamental
thing
of
the
triage
is,
is
working.
D
D
A
D
We
may
need
to
not
start
looking
at
some
of
the
other
ones
now
and
looking
at
issues
raised
on
those,
but
I
think
we
need
to
get
structured
a
little
bit
to
make
that
work,
meaning
we
had
the
idea
of
a
like.
Oh
you
just
choose
which
ones
you
fans
have
been
interested.
We
may
actually
need
to
get
together
the
active
people
and
say
no.
These
are
the
ones
that
actually
are
active
and
we
should
look
at
because
there
are
just
so
many
of
them.
Yeah.
B
D
A
D
D
D
A
D
A
A
That
sounds
good,
so
I
just
wrote
as
the
agenda
a
thing
Glenn
will
reach
out
to
Jewish
and
West's
Jerry,
schwas
and
John
to
see
to
see
if
we
can
set
up
some
sort
of
interlock,
possibly
in
the
expressed
EC
meeting,
mm-hmm
sound
good
and
then
issue
schedules
goes
around
when
to
close,
because
I
assume
you
guys,
you
know
you
can
close
now
right.
We.
D
Can
but
mostly
I'm
asking
I,
don't
like
to
be
presumptive
about
what
I'm
a
man
and
not
lucky,
so
you
should
asking
the
initiator.
Are
you
ok
to
close
this
and
they
just
close
it?
We've
answered
a
question:
they
missed
a
slash.
Our
yep
Express
is
still
working
I'm
good
with
that,
but
we
need
to
have
some
sort
of
rules
between
us
about
when
we
go
in
and
I.
Think
like
that
things
hanging
around
for
more
information
for
a
month
will
be
cut
a
lot
of
them
down.
You
know
pretty
yep.
A
D
A
A
D
G
Ahead
in
the
last
meeting
waste
motion,
there
were
some
issues
with
the
timeouts
for
the
Express
status
board.
Yes,
I
often
take
some
time
to
try
and
look
at
that.
Try
and
see
if
I
can
improve
the
timeouts
by
changing
the
queries
from
using
graphic
up
instead
of
using
the
REST
API.
So
I
should
create
a
working
progress
PR
on
that
Claire,
soon
open
up
the
pkgs.
Oh,
that.
A
A
A
A
C
C
C
A
A
C
F
I
think
we
they
have
and
they
are
open
to
change
the
default
from
current
LTS
and
I'm,
not
sure
where
that
landed
and
I
think.
Maybe
we
need
to
revisit
that
so
that
they
start
off
with
deltas
as
the
default
for
new
projects
and
and
I
know
that
we
sort
of
there
was
some
discussion
around
starting
off
with
the
oldest
LTS
at
the
moment,
which
is
debatable,
but
as
long
as
they
change
out
yes
to
change,
current
LTS
is
probably
valuable,
regardless
of
which
ideas
and
the
rest
we
can
handle
with
the
custom
configs.
F
F
C
A
F
A
B
A
F
A
Paul
get
my
support.
Levels
against
were
yeah
blog
post
on
adding
to
yeah
okay,
so
those
are
all
I
think
once
we
get
through
more
the
next
step
of
the
support
levels,
document
pave
paths
for
minimizing
size
of
published
packages
and
maximizing
the
availability
of
the
content,
tests,
etc.
I
think
we
did
have
at
least
one
content
was
developed
around.
A
If
we
have
those
as
templates,
then
those
are
the
things
we
might
want
to
help
get
into
projects.
So
I
think
it
leave
that
to
lows:
define
baseline
practice,
practices
for
testing
baseline
practices
maintain
errs
key
packages
for
open,
enrollment,
best
practice,
managing
LTS
and
release
lines
for
user
land
modules.