►
From YouTube: 20190628 sig testing commons
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
It's
been
sort
of
a
just
forever
forward
thing
at
the
moment,
but
if
we
set
where
the
finish
line
is
then
maybe
I
don't
know
if
everything
can
be
consolidated
into
a
giant
host,
but
Claudia
is
doing
a
really
great
job
of
doing
you
know
four
or
five
images
at
a
time.
I,
don't
he's
done
three
parts
right
now,
part
three
is
up
I,
don't
know
how
few
that
hits
us,
but
I
will,
though
he
knocks
I
didn't
for
me.
A
B
B
B
Don't
I
think
for
most
people
that
actually
run
the
conformance
tests
in
the
wild.
They
don't
care
right
they
if
it
takes
a
while
it's
a
one-shot
deal.
It
is
not
a
something
that
they
do
continuously
I.
Think
for
test
automation.
It
does.
It
is
a
bigger
deal
and
you
know
if
they
want
to
be
able
to
spin
up
faster
ci
runs
paralyzation
matters.
B
I
think
we'd
have
to
actually
check
inside
of
the
tests
to
make
sure
there
aren't
collisions
on
things
like,
for
example,
when
we
have
parallelization,
we
do
everything
inside
of
a
namespace
scoped
items.
There
is
no
check
for
collision
and
namespace
scoping
on
some
of
these
things
and
they
do
collide
and
we've
seen
it
before
really.
A
B
C
B
The
amount
of
ROI
on
it
it
depends
upon
what
you're
doing
right
from
an
end
customer
perspective
of
just
being
able
to
run
the
tests.
It's
not
don't
care
mm-hm,
they
look
care
if
they're
running
it
multiple
times
so
I'm
not
I,
think
the
priority.
This
is
like
I
consider
this
backlog
and
they
take
it
off.
116,
okay,.
A
Yeah,
the
when
I
did
an
investigation
into
this.
The
one
thing
I
found
was:
we
only
have
one
test
that
is
currently
marked
conformance
and
Syria,
but
the
reason
for
that
I
put
at
the
very
bottom-
it's
a
PR
from
like
three
years
ago
and
it
seemed
kind
of
tenuous
that
it
needed
to
be
cereal.
It
wasn't
like
disruptive,
it
was
just
saying,
like
we
try
and
make
a
race
by
doing
lots
of
things
and,
like
I,
didn't
know
why
that
had
to
be
cereal,
so
I
had
proposed
at
least
maybe
four.
A
B
B
A
B
A
D
A
A
D
One
thing
that
might
be
helpful
is
like
just
getting
the
skeleton
import
restrictions
file
just
in
the
repository,
and
then
we
can
infer
mentally,
like
figure
out
what
imports
imports
that
we
should
prune
and
not
and
follow-up.
Yours
yeah.
A
A
Is
slightly
in
front
of
us
just
because
this
mentions
cube
test
I
had
just
been
looking
at
some
of
that
stuff.
The
other
day
and
I
noticed
that
there's
the
cube
test
and
there's
cube
test
too.
Does
anyone
know
the
status
of
that?
I
saw
like
what
looked
to
be
a
bunch
of
em's
but
as
cube
test
to
the
norm
for
other
folks
now,
I
don't
use
either,
but
I'm
just
curious.
No.
B
Cue
test
two
was
supposed
to
be
like
a
pruning
and
refactor.
Instead
of
the
testing
framework
and
I
think
it's
last
I
heard
it
was
not
actively
being
worked
on.
Okay
I
got
a
hard
stop
at
ten,
but
so
let's
try
and
get
through
ten
my
time
so,
whatever
your
time
I
got
hard
stuff
at
the
end
an
hour
a
lesson
today,
because
I
have
a
meeting
overlap.
B
E
E
B
So
the
developer
track
is
useful.
If
you're
trying
to
do
a
call
to
arms,
that's
usually
the
purpose
of
the
developer
track,
so
I
think
if
we
do
that,
we
should
talk
about.
You
know
why
we're
doing
this
and
try
to
solicit
feedback
and
try
to
get
more
people
engaged
I.
Think
that
is
a
worthwhile
effort.
E
B
We
would
have
to
be
like
they're,
either
insurers
or
deep
dives,
I.
Think
I
think
this
apana
be
like
one
of
those
smash
stocks.
I
think
they're
kind
of
doing
that
nowadays,
where
they
have
they
used
to
have
two
parts
where
they
had
insurance
each
time,
and
now
they
have
the
smash
one
which
is
basically
like
just
put
them
together.
It
seems
like
this
would
be
one
of
those
like
put
them
together.
Talk
about
testing
in
communities
and
we
might
have
to
open
ourselves
up
to
a
broader
set
of
details.
B
B
D
I
see
two
different
angles
for
this
potential
talk
around
testing
comments
yeah
the
first
one
is,
is
what
you
mentioned:
Tim
like
trying
to
get
more
developers
to
help
by
talking
about
like
some
of
the
details
and
challenges
we're
facing,
but
then
the
second
one
is
like
benefits.
System
like
by
providing
an
external
framework,
like
other
people,
can
write
their
own
custom
tests
for
their
third-party
plugins
or
whatever
not.
C
D
B
A
A
So
I've
had
fuse
yeah
Jordan
Liggett
commenting,
but
there
there's
just
sort
of
an
open
question
as
to
there's
these
functions.
That
will
do
this
waiting
and
the
question
is:
do
we
want
to
provide
an
opt-out
just
at
the
start
of
the
suite
or
internal
to
that
logic,
because
there's
certain
tests,
that
will
say,
for
instance,
like
destroy
a
node,
now
wait
for
all
nodes
to
be
ready
so
like
do
we
need
to
be
skipping
all
that
logic
as
well?
A
A
So
I
hits
I
had
started
doing
that
Jordan
actually
had
a
PR
that
did
pretty
much
the
same
thing
so
I
deferred
to
him.
Then
I
got
put
on
his
back
burner,
so
I
picked
it
up
again
and
he
had
said
he
had
iterated
and
had
conversations
and
liked
this
solution
of
simplifying
some
of
that
logic
in
not
making
some
special
case
from
master
labels
and
in
taints
and
stuff
and
then
providing
some
way
to
just
opt-out
of
the
weight.
A
B
A
B
Right
so
I
think
that's
all
we
got
for
today.
I
will
update
the
meeting,
invite
as
well
as
the
community
facing
stuff
and
if
folks,
wanna,
asynchronously
I
might
a
certain
ously
do
this
today,
I
can
walk
through
the
rest
of
the
backlog
and
try
to
get
a
little
more
room.
So
next
time
we
chat,
we
can
have
a
little
bit
of
a
discussion
about
the
backlog,
there's
only
one
page,
so
there's
whatever
in
many
issues
in
one
I'll
take
care
of
it
and
Gentile.