►
From YouTube: 2019 09 09 Memory Team Meeting
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
You're
basing
this
green
eye
all
right,
so
past
meetings,
we've
just
been
basically
reading
through
the
agenda
on
a
weekly
meeting
and
haven't
really
gone
through
the
boards
and
everything
so
we're
gonna
change
this
format
a
bit
can
still
have
announcements.
Topics
we
want
to
cover.
People
can
contribute
to
the
agenda
and
then
near
the
end
of
the
meeting
or
halfway
through
the
meeting,
we'll
jump
over
to
the
board
and
make
sure
that
we
covered
in
topics.
People
need
to
cover,
go
through
work
in
progress,
make
sure
we're
not
blocked
on
anything
answer.
B
A
That's
it
for
the
agenda
change,
so
those
four
topics
got
the
retro
tomorrow
early
morning
for
us
folks,
it's
been
glorious,
that's
2:00
in
the
morning,
if
you
wanna,
join
otherwise
its
normal
time
for
pretty
much
everyone
else
links
there.
It's
on
my
count,
I,
don't
think
I
invited
Euler
aside
just
to
see
where
can
it
go,
but
I
can
invite
you
if
you
want
and
then
the
automatic
acing
team
retro
was
created
this
morning,
so
we
started
thinking
about
12
3.
A
You
know
we're
about
we're
at
the
midpoint
I
guess
the
last
build
or
the
I
guess.
The
candidate
build
will
be
cut
on
the
17th
for
12
3
and
then
milestone
officially
closes
on
23rd
so
about
how
much
time
we
have
left
for
this
milestone.
So
any
topics
anything
else
folks
want
to
cover
on
updates.
Before
we
go
over
to
the
boards,
you
all
right.
So
we
have
a
couple
boards.
A
A
A
Next
thing
up
would
be
staging
and
I.
Don't
I,
don't
know
if
we
have
issues
other
than
enabling
it
on
staging.
It's
not
a
whole
lot
more
data
on
their
new
e.
Is
it
the
same
process?
We
asked
you
out,
we
give
jar
of
the
configuration
and
we
ask
him
to
push
it
out,
staging
pretty
mouth
this,
okay
and
then
how
much
time
do
we
want?
Oh
well,
we
got
the
chronic
worker.
Is
that
gonna
stop
us
from
moving
to
staging
you
think.
B
So
I
think
for
staging.
We
would
go
ahead
even
like
when
this
deficiency,
the
female,
but
for
the
canary
I,
think
we
need
to
solve
chronic
first,
okay,
so
I
guess
like
as
soon
as
I
come
out
and
close.
This
reconfigure
them
get
up.
The
dark,
I
just
gonna,
leave
a
comment
on
this
issue
that
we
should
just
go
ahead
and
enabled
on
staging
knowing
that
these
are
the
deficiencies,
and
this
is
the
data
that
we
are
looking
for.
A
A
Can
add
some
comments
to
you
in
the
end
of
the
day
not
getting
to
it?
So
so
it's
not
looking
like
sorry,
someone
talking
to
religious
me,
I
going,
must
been
nako
all
right.
So
it's
not
looking
like
we're.
Gonna
get
Puma
into
production.
Probably
this
milestone.
Unless
we
get
the
chronic
issue
fixed,
very
quickly
right
or
even
give
it
sometimes
it.
A
So
we'll
need
to
start
communicating
that
out,
because
one
of
the
goals
for
twelve
through
is
to
get
poom
out
in
the
production.
It's
fine.
We
had
a
lot
of
interruptions
and
distractions
this
milestone.
So
it's
not
a
big
deal.
It's
just
better
to
communicate
it
earlier.
We're
not
gonna
hit
that
milestone,
so
I'll
work
with
Larissa
and
product
to
figure
out
how
to
communicate
that
up
see
other
things.
A
So
we
move
some
of
the
research
for
cpu-intensive
endpoints
out
of
12
3,
because
the
chronic
duration
issue
came
into
three
and
was
assigned
or
Alexei
Alexei
was
assigned.
Although
CPU
memory
intensive
inflate
request,
so
we
unassigned
them
move
them
off
to
twelve.
Four
will
consider
them
for
the
next
milestone.
B
A
B
B
B
Gonna
finish
just
one
aspect
out
of
that:
it's
out
of
the
related
tissues,
I'm
gonna
only
managed
to
finish
the
combo
site
status,
the
one
that
it's
open
there
is
like.
There
is
on
the
bellow
of
the
items
to
finish.
If
you
could
scroll
down,
I'm
gonna
updated
with
the
status
but
technically
I'm
not
gonna,
be
able
to
finish
pipe
writing
process
worker.
This
release.
I'm
gonna,
improve
the
composite
status
than
diagonal
methods
like
this
week,
but
the
compass
are
like
the
Python
process.
Worker
has
to
be
done
in
the
next,
so
I'm
gonna
finished.
A
B
So,
actually,
like
we
have
this
finished
15
51
52,
one
site
that
does
the
never
saw
a
decision
like
I'm
gonna
close
this
issue
with
the
magic
West
and
cross
link.
This
issue
because,
like
we
still
him
and
self,
like
the
all
aspects
of
these
simple
pipeline
service,
so
like
we
gonna,
have
to
move
the
to
next
milestone.
B
But
commenting
like
what
pre
our
work
we
did
because,
like
all
of
the
aspect,
needs
to
be
done
and
I
think
he
like
this
issue,
the
way
how
it's
very
constructive,
it's
kind
of
like
the
big
picture,
but
pretty
much
like
the
point
number
two
and
number
three.
Once
we
start
working
on
them,
they
should
be
like
the
separate
issue
that
we
will
be
closing
that
way,
meaning
for
this
bigger
story.
B
B
D
Yeah
I
think
I
kind
of
left
that
open
saying
if
we
really
think
it's
needed,
I
can
explore
it,
and
my
testing
I
haven't
really
seen
any
limits.
Beam
hits
only
said,
one
testicle
higher
sequel
queries
and
those
the
sixteen
friend
I
guess.
The
question
is:
if
we
really
wanted
and
yes
we
should,
we
can
open
up
narration,
look
at
exploring
or
to
tester.
B
So
I
guess
it's
really
like
about
like,
including
our
testing.
What
we
search
now
propose
how
to
configure
like
stating
canary
and
github.com
Oman,
because
right,
because
the
different
configurations
and
I
think
for
us
to
like
across
this
issue,
like
is
to
have
some
initial
recommendation
on
how
to
use
from
Houma.
C
D
Do
you
think
it's
blocking
this
issue
for
me
close
down
yeah,
so
I
think
it's
a
good.
Do
you
think
it's
blocking
us
from
being
like
I,
think
we're
all
kind
of
all,
but
all
been
officially
saying
that
workers
for
worker
for
friends
it's
the
way
to
go
at
least
to
begin
with?
Is
that
without
testing
be
required
to
recommend
a--,
or
do
you
think
that
we
can't
recommend?
Yet
until
we
know
this
and
the
testing
results
I.
D
D
D
Cisnormative
mood
phrasing
is
that
respect
in
there
to
be
like
a
really
big
secrets
kind
of
issue
in
there.
What
DB
connection
limits
that
with
block
us
from
moving
ahead
of
for
for
but
I
guess
for
calm,
for
this
movie
should
probably
just
I'm
not
actually
looked
into
her
to
configure
eleven
had
to
do
a
DB,
Louis,
I'm
hoping
it
would
be
just
an
omnibus
and
assessing
yeah.
A
D
D
Although
looking
at
that
issue,
there
is,
it
seems
to
suggest
that
there's
some
kind
of
automation
configuration
that
we
sidekick,
but
not
a
unicorn.
So
doesn't
it
be
anything
in
place
to
scale
up
the
cool
connection
tool
depending
how
many
threads
we
have
toll
for
Puma,
yeah
I.
Think
that's
what
Stan
was
asking
that
this
year,
how
we're
gonna
handle
that
that's
a
good
question.
D
Okay,
that
their
thing
is
that
something
is
being
checked
in
I
guess
a
paper.
We
should
be
recommending
the
mirror
threads,
that
we
have
that
this
app
would
have
for
Puma
to
make
it
match
psychic
all
the
way.
It's
not
automated,
yeah,
okay,
I
will
do
a
test
with
that.
Then
I'll
do
a
four
four
tests
with
the
four
with
a
16-2
limiter
and
see
how
the
Jax
but
yeah.
D
D
A
D
B
D
I
personally
as
position
here
where
I've
had
to
like,
make
an
official
recommendation
that
would
be
actually
putting
dogs,
yeah,
I've,
always
kind
of
pushed
back
from
that
saying.
When
these
people
who
know
you
know
get
lob
more
than
someone
who's
really
been
here
for
a
couple
months
to
be
able
to
say
definitively
that
these
are
the
right
things.
D
The
tanking
architecture,
for
example,
is
don't
like
that
by
the
Emperor
team
and
we
kind
of
words
them
on
that
that
he
says
I
mean
we
are
starting
to
drive
that
more
of
something
appointed
speeds
italia,
but
you
want
to
come
back
against
and
give
the
actual
recommendation.
I
will
be
confident
unless
someone
who's,
you
know
an
expert
or
a
stake
holder
and
Gettler
can
actually
come
back
so
yeah
that
sentence
writers
as
like
a
good
way
forward.
B
Because
for
me,
it
seems
like
that
the
course
of
action
to
resolving
this
issue
in
context
of
10k
reference
architecture
is
like
it's
get,
Punk
a
reference
architecture
and
test.
We
put
our
neighborhood
on
different
configurations
like
threads
and
put
my
memory
killer
and
validate
what
values
you
get
I.
B
Think,
like
we
don't
have
anything
else
that
would
allow
us
to
cause
this
issue
and
pretty
much
the
same
process.
We
have
to
be
redone
for
every
other
architecture
that,
like
we
gonna,
be
recommending
because,
like
what
we
are
now
doing
before
mice
like
we
are
ranked
on,
it
will
come
on
production,
but
pretty
much
like
we
should
have
already
the
open
issue
to
a
label
Puma
for
everyone,
so
also
like
for
our
users
are
gonna.
B
D
Not
I
still
need
some
we're
at
the
stage
right
now,
where
we're
kind
of
starting
to
hope
to
take
on
the
tank
environment
as
well
as
offering
future
reference
environments,
and
we
would
be
driving
that
we're
still
very
much
in
the
early
days
that
I'm
proving
how
to
do
that
in
an
automated
fashion.
The
tanky
environm
annually
by
the
infantry
and
certainly
I,
can
ask
them
for
teens
in
these
pivoted
to
Pima.
No,
and
certainly
future
reference
of
areas
will
be
improved
by
default.
I
would
probably
consider
a
unicorn
record.
D
We
can
ask
the
team
stood
up
now,
or
can
we
be
trying
to
figure
out
how
to
get
access
to
the
environment?
Ursula
backs
pivot
and
contesta?
That's
the
kind
of
options
we
have
at
the
moment
in
the
future.
Hopefully,
we'll
have
the
tanky
environment
and
our
office
environments
and
an
automated
fashion,
and
we
can
hopefully
be
able
to
pivot
it
quicker
to
different
France
workers,
configurations
etc.
D
A
No,
it
seems
like
there's
multiple
different
areas
that
we
need
to
define
the
best
default
values
right.
We're
gonna
have
the
10k
architecture,
we're
gonna,
have
the
production
environment.
We're
gonna,
have
recommendations
for
self-managed
I
would
assume
so
seems
like
this
issue
needs
be
broken
down
a
bit
more
and
have
some
specific
targets
for
where
we're
defining
the
best
values.
So
I
will
kick
it
out
to
12,
for
when
we
can
work
on
breaking
it
down
further
and
maybe
get
some
feedback
from
product
on
where
we
need
to
define
the
best
values
and
what's
position.
A
B
I
mean
kind
of
like
asking
quality
team
to
take
ownership
of
the
10k
puma
testing,
because
I
think
that
quality
team
has
all
the
tools
and
work,
on
the
other
hand,
was
I,
could
ruin
the
infrastructure
team
on
the
settings
for
the
github.com.
Because-
and
my
me
with
having
gonna,
see
some
pattern
and
I
have
the
half
days
and
I
put
some
universe
are
defaults
because
we
like
I,
think
that
what
we
miss
right
now
for
the
Puma.
B
We
have
a
ground
for
naming
that
on
the
production.
But
we
missed
the
issue
for
and
I'm
doing
that
for
everyone
by
default,
and
we
also
miss
the
steps
for
and
I
think
that
by
default
for
everyone,
because
as
soon
as
we
may
put
it
on
the
production
like
it
shall
be
default
for
Omnibus.
But
no
night
before
everyone
assumed
impossible.
D
Yeah,
it's
just
a
pink,
pink
and
I
will
be
more
than
happy
to
drive.
Testing
yes,
take
a
vibrant
with
FEMA.
That's
doesn't
as
you've
raised.
It's
just
also
at
the
timing
was
to
make
a
surrender.
If
you
want
to
raise
an
issue.
I
says
we
wanted
benchmark,
come
up
against
the
tanky
environment
with
various
configurations
of
Puma.
A
Going
back
to
what
you
said
Camille
about
this
issue,
maybe
we
just
retitled
this
to
excuse
me:
define
the
best
default
values
in
omnibus
for
Puma.
Consider
and
that'll
make
this
issue
more
specific,
and
then
it
sounds
like
we
need
a
another
issue
to
actually
turn
it
on
by
default
within
omnibus
as
well.
Right,
yes
and.
B
Pretty
much
manganese,
another
issues
will
be
blocked
by
this
one
I
guess
it
goes
like
this
should
have
some
sign
divorce
before
we
enabled
it
for
everyone:
yep,
okay,
I'm
I'm,
not
sorry
worrying
about
reg
the
correlation
between,
like
this
related
issues,
the
door
of
the
table
that
we
discussed,
because
there's
like
this
butter,
my
configuration
of
Puma
server
for
performance
benchmark.
If
you
could
go
back
to
this
previous
page,
we
actually
asked
out
of
the
time
almost
I
think
because
it
was
for
this
performance
but
mark
people
top
Capital
performance,
baseline
I.