►
From YouTube: 2020 08 31 Memory Team Weekly
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
All
right
welcome
to
the
memory
team
weekly
meeting.
It
is
august,
31st
jump
right
into
it.
So
from
the
retro
standpoint,
the
retro
recording
is
gonna
happen
here
in
a
couple
days,
so
still
need
feedback
from
alexi
nicola
and
camille,
and
I
will
I
will
follow
up
with
nikola
when
he
gets
back
tomorrow.
A
In
and
review
the
okrs,
I
figured
we'd
run
through
them
real,
quick
here.
The
first
one
in
our
list
is
the
70
sedu
ratio.
Does
anybody
have
questions
on
what
that
actually
means?
A
The
quick
overview
is
we
hit
70
percent
of
what
we
say
we
are
going
to
deliver
in
a
milestone,
and
one
thing
that
we
need
to
do
is
come
to
an
agreement
on
when
we
mark
items
as
deliverables.
I
think
one
last
milestone.
We
went
through
at
the
beginning
of
the
milestone
and
marked
items
as
deliverables,
but
we
we
have
not
done
this
this
milestone,
so
I
figured
we
could
do
that
right
now.
A
So
if
we
run
through
them
real
quick,
not
all
of
these
are
assigned,
but
this
one
was
just
added
this
milestone.
Camille.
I
don't
know
if
you're
there,
but
will
this
be
a
deliverable
for
this
milestone.
A
B
A
Cool,
thank
you,
dynamic
image,
resizing
proof
of
concept.
We're
planning
on
closing
this
one
out
this
milestone,
correct,
alex
needs
to
get
your
name
on
it.
It's
yours.
C
Sorry
because,
if
please
repeat
that,
because
it
distracted
yes,
this
one
you're
asking
if
it
will
be
delivered
in
this
milestone.
Yes,
I
think
it
should,
because
I
have
a
feature
partner
request
on
review
and
after
that
we
are
ready
to
run
some.
A
Okay,
futuregate
matthias.
I
guess
it's
a
sign
to
like.
C
A
Okay
looks
like
you
opened
up
this
one
like.
C
This
one
I
put
it
on
hold
because
we
are
not
sure
about
gm-
is,
is
being
used
in
our
solution.
So
actually
I
checked
it
it's
not
available.
So,
technically
speaking,
we
could
even
close
it
and
reopen
a
new
issue
to
ask
for
edit
the
home
charts
for
our
home
folks.
D
D
D
D
C
B
Yes,
so
this
I
mean
they're
all
out
ones
right.
The
videos
are
skip
time
kind
of
need
to
remove
the
feature
clock
for
these
features.
Like
the
issue
is
there
because,
like
I
didn't
uncheck,
I
I
didn't
check
the
last
items,
which
is
like
removal
from
the
codebase
okay,
I'm
kind
of
waiting
a
few
more
days
to
ensure
that
there
is
nothing
wrong
but
except
the
first
report
from
some
person
saying
that
the
behavior
changed.
But
I
agree
that
this
is
the
right
behavior.
There
was
no
further
reports
on
that.
B
A
Okay,
let's
see
image
scaling
mathias,
you
get
the
next
two.
D
It's
really
tricky
with
this
stuff
because
we
don't
know
if
we're
even
going
to
ship
this
right.
So
I
don't.
B
A
Okay,
that
makes
sense,
then
matthias
you're
probably
closest
to
this
one.
D
Yeah
this
is
yeah.
This
is
actually
going
going
well,
like
chinu
did
some
really
good
sleuthing
there
and
he
found.
I
think
he
has
a
pretty
good
lead
on
what
might
be
causing
this,
and
he
has
a
proposal
as
well
for
how
to
fix
it,
and
it
also
got
just
approved
by
security.
So
it
looks
like
we
might
be
getting
a
fix
out
for
this
for
30.4,
so
I
think
I
think
it
I
think,
we'll
yeah.
We
will
try
to
ship
a
fix
with
34.
D
Yeah,
so
I
I
kind
of
need
to
redo
that,
but
I
still
want
to
do
it
in
this
milestone.
It
shouldn't
be
too
hard
because
of
because
we
changed
the
way
we
integrate
with
graphics
magic.
We
need
to
do
it
in
a
different
way
now,
but
yeah.
I
think
we
can
do
it.
A
Oh
dang,
I
think,
is
this
one:
gonna
make
it
in
13-4.
D
I
think
we'll
I
have
my
doubts.
I
mean
we.
B
D
A
Okay,
this
isn't
even
within
our
group
that
one's
probably
not
going
to
happen.
That's
not
a
deliverable,
probably
not
going
to
happen.
What
about
this
last
one
image
resizing.
C
No,
I
don't
know
about
it,
I
think
it.
Maybe
we
keep
like
pushing
this
issue
every
discussion,
so
I
think
we
either
need
to
close
it
or
maybe
rephrase
it
or
maybe
even
to
move
to
the
next
milestone,
because
initially
it
was
like
in
a
planning
bunch.
So
when
we
try
to
lay
out
the
possible
things
we
need
to
learn
about
initiating,
but
it
seems
like
we
need
to
push
it
forward
for
now.
A
One
right
now
yeah
this
one's
not
going
to
make
it.
A
A
A
All
right
so
back
to
okr
so
convert
to
training
maintainers.
So
I'm
talking
with
folks
individually
on
this
one.
I
have
two
definite
plans
from
two
folks
on
where
they
are
and
the
other
two.
I
need
to
talk
to
you
about
their
timeline
so
and
when
we
talk,
I'm
really
it's
it's
all
about
your
comfort
level,
I'm
not
trying
to
force
anyone
into
converting
to
maintainer
this
quarter
or
any
preconceived
timeline
I
may
have
so
if
it
feels
like
I'm
pressuring
you
please
I'm
not.
I
just
want
to
know.
A
A
Let's
see
next
one
implement
two
memory
performance
related
improvements,
so
image
processing
is
progressing
when
this
percentage
here
is
what
the
progress
that
we
need
to
indicate
how
far
we
are
along
on
the
okr,
and
I
just
haven't
updated
it.
Yet,
I'm
not
sure
what
I
would
put
there
70
would
indicate
successfully
done
with
that
okr
and
it
just
needs
to
be
accurate
at
the
end
of
the
quarter.
If
we
did
not
get
to
them.
A
If
we
don't
successfully
close
out
the
lkr,
we
just
explain
why,
like,
if
we
got
distracted
by
something
else
or
prioritized
on
something
else,
I
will
update
it
in
the
kr,
but
we're
making
the
right
progress
here.
We're
progressing
on
image
processing
and
I
think
nikola
switched
over
to
cash
sql
calls.
Last
week
and
shin
you
will
start
contributing
to
that
once
you've
done
the
telemetry
work.
A
A
The
roadmap
is
out
there
and
I
have
a
link
at
the
bottom
of
the
page
and
the
other
one.
I
was
going
to
work
on
an
example
of
like
decision
logs
like
some
of
the
bigger
projects
that
we
worked
on,
I'm
not
sure
how
I'm
going
to
do
it.
Yet.
I
think
I'll,
probably
just
do
a
brief
outline
linking
back
to
mrs
and
issues
that
we've
done
in
the
past.
So
it's
not
basically
rewriting
all
the
content
of
the
issues
that
we
already
have.
A
A
Jump
over
to
progress,
update
from
a
feature
flag
standpoint
after
talking
to
camille
and
some
other
folks,
I've
gotten
involved
to
help
organize
finishing
this
up
and
hand
off
long-term
ownership.
It's
not
something
that
should
stick
with
memory
team,
something
that
camille's
working
on
right
now
but
doesn't
fall
into
our
charter.
It
seems
like
a
progressive,
I'm
sorry.
The
delivery
team
would
be
the
most
likely
owner
going
forward
after
talking
to
a
couple
managers
so,
and
then
camille
had
to
comment
about
be
largely
complete
in
this
milestone.
Chameleon
and
verbalize.
B
B
B
And
which
I
kind
of
asked,
if
we
could
do
zip,
because
it
opens
a
few
interesting
opportunities
and
the
second
one
like
there
was
like
discussion
about
the
dogskytrap.com,
which
is
using
the
pages
under
a
discussion
about
moving
dogs
gitlab.com
out
of
the
github
pages,
where
I
try
to
convince
people
that
it's
bad
idea
and
we
should
rather
look
for
a
person
external
to
the
pages,
seem
to
implement
some
features.
So
we
actually
got
candidate
that
did
go
very
far
into
implementing
one
of
the
missing
features.
B
B
It
was
one
of
the
two
requirements
for
that.
The
dogs
get
club
team
do
require
for
the
gitrup.com
second
one,
which
is
like
multi
pages
support
per
project.
It's
gonna
be
actually
possible
with
the
object,
started.
Zip
architecture
so
like
these
two
tracks
really
like
long
time,
should
solve
the
the
dock
github.com
deficiency
that
they
are
not
really
the
efficiency
that
the
problems
that
they
are
struggling
with.
B
Quick
smile,
if
you
are
interested
about
the
threat
there
is
this
issue
with
a
note.
You
can
read
more
about
like
the
course
of
action
that
did
happen
and
gitlab
runner,
I'm
still
like
runner
maintainer.
B
So
I'm
kind
of
started
reviewing
some
of
the
runner,
mrs
about
after
script
and
cancelling
state
support
trying
to
help
the
people
right
move
that
along
because
it
was
something
that
was
kind
of
stupid
for
like
probably
half
a
year
or
even
more
already,
so
my
involvement
there
is
like
pretty
little,
but
it
still
takes
some
time
to
do
these
reviews
and
go
through
this
poc
of
cancer.
Instead
that
I
added
that's
for
me,
okay,.
A
D
It
under
the
does
do
and
was
it
a
seido
ratio
anyway
yeah.
So
we
have
a
working
implementation
that
uses
second,
as
the
security
team
requested,
so
which
is
like
a
binary
that
we
own
and
build
and
it
uses
lip
graphics
magic,
so
not
the
cli
tool.
D
So
we
don't
shell
out
to
graphics
magic
anymore,
but
we
embed
the
library
with
our
own
binary
that
we
build
so
so
it's
working
there's
a
couple
open
issues
like,
for
instance,
we
we
need
to
make
sure
it
runs
everywhere
first,
so
it's
like
the
build,
isn't
even
working
right
now,
just
me
to
make
these
libraries
available
and
also
there's
the
question
of
mac
os,
which
is
not
officially
supported
platform,
but
a
lot
of
developers
use
it,
so
they
wouldn't
even
be
able
to
build
workhorse
anymore.
D
D
I
can
you
edit
this
yeah.
I
think
I
mentioned
that
one
earlier
yeah,
so
that
was
about
like
looking
into
why
all
these
requests
the
prometheus
fail,
and
there
is
some
indication
that
this
might
be
due
to
ssl
handshake
failures
or
like
we.
We
do
not
expect.
We
never
built
this
with
ssl
in
my
https
in
mind.
D
So
apparently
a
much
larger
group
of
customers
than
we
anticipated
are
running
from.
If
we
can
turn
the
embedded
prometheus
behind
an
like
an
ssl
proxy
like
nginx,
or
something
like
that
so
yeah.
So.
D
Discussion
started
about
like
yes,
certificate
management
and
all
that
stuff,
and
it
looks
like
a
simple
option
might
be
to
just
disable
these
checks
yeah,
because
there's
no
like
sensitive
data
already
that
crosses
the
wire
there,
so
it
should
be
okay,
multi-node.
E
A
E
Yep
yeah
so
sid
feel
strongly
about
this.
It
feels
like
we
are
being
disrupted
from
below
by
getty,
primarily
and
would
like
to
see
us
get
the
two
gigs.
I
did
express
that.
I
think
the
comments
you
hear
about
git
lab
on
hacker
news
and
places
that
we
are
bloated
and
you
know
too
heavy-
are
likely
talking
about
memory,
consumption
and
performance
and
package
size.
So
I
think
you
kind
of
need
to
address
all
three
which
we're
working
on
to
really
kind
of
address
the
challenge
of
get
to
something
from
below.
E
But
you
know,
I
think,
if
we
can
start
making
some
progress
after
we
work
on
the
imagery
sizing.
That
would
be
good
if
we
have,
if
we
find
some
other
big
opportunity
to
improve
performance,
I'm
happy
to
rave
out
the
that
prioritization
and
and
get
sid
for
his
buy-in
on
that
relation.
But
for
now
it's
he's
on
board
with
finishing
image,
resizing
and
then
picking
up
the
ruby
two
f7
and
the
and
the
pre-forking
for
the
four
gaming
before
sorry
preloading
before
forking
worked
next.
E
Github
that
ruby
207
improvements
have
a
performance
improvements,
embed
them
as
well.
I
was
flipping
through
some
of
the
ruby
3x3
stuff,
where
you
want
to
make
it
three
times
faster
than
and
ruby3.0,
so
so
we'll
see,
but
I
imagine
we'll
get
some
performance
benefits
as
well,
along
with
the
rubik's
27
stuff.
C
E
E
Yeah
I
mean
we'll
see
we'll
see
where
we're
at,
but
15
7
would
be.
What
is
that
november
december?
I
think
13
of
4
is
september,
5
october
6
november
yeah
december.
So
the
end
of
year.
A
Okay,
that's
it
for
what's
on
the
agenda,
any
other
topics
anybody
wants
to
cover
today.