►
From YouTube: 2020-08-04 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
B
Hi
welcome
this
is
alita.
Were
we
just
waiting
for
folks.
B
A
B
A
Also,
let's
probably
let's
start,
I
would
like
first
to
discuss
one
one
thing
that
I
I
know
most
likely
it's
already
confirmed
is
the
issue
731
from
the
specification,
and
I
think
already
it's
assigned
to
you.
Are
you
planning
to
to
do
a
change
to
the
specification
to
to
support
that.
B
So
here,
specifically
bogdan
the
if
this
did
come
up
in
the
morning
and
in
the
spec
meeting,
and
the
idea
was
that
you
know
our
team
and
some
of
my
engineers,
as
you
know,
have
been
working
on
verifying
the
spec
as
well
as
kind
of
handling
cumulative
metrics
for
cortex
and
export
and
prometheus,
and
and
so
we
have
done
a.
You
know,
fair
bit
of
analysis
that
we
can
certainly
submit
for
in
support
of
what
needs
to
be
changed
in
the
spec.
B
But
ultimately,
you
and
josh
would
obviously
have
to
review
it
before
anything's
merged
right.
So
that
was.
That
was
the
reason
why
you
know
my
name's
there
as
one
of
the
okay.
A
A
Okay,
then,
the
next
thing
that
I
would
like
to
discuss
is
the
seven
to
five.
I
know
I
did
not
get
any
feedback,
but
I
would
like
to
know
what
are
what
do
people
think
about
it,
and
I
know
datadog
has
some
some.
They
promised
to
me
that
they
comment,
but
they
did
not.
A
So
if
anyone
has
any
opinion
on
that
one,
I
will
post
it
here
in
the
chat.
Please
please
comment,
because
I
would
like
to
resolve
it
before
ga
to
call
a
decision
on
this,
and
I
would
like
to
hear
others
opinion
on
this
as
well.
B
A
Sooner
sooner
is
better
anytime,
yeah,
okay,
so
these
are
the
two
main
issues
that
we
have
in
the
specs.
I
think-
and
there
is
another
one-
it's
I
think
three
six
is
the
one.
A
A
Yes,
it's
six
one,
seven,
six,
one
seven
here
on
this
one!
I
also
need
to
know
what
people
believe
if,
if
we
believe
it's
important
to
support
this-
and
we
need
this-
we
just-
I
just
need
to
make
a
decision
on
this
one
and,
and
so
far
so
far
I
haven't.
I
haven't
heard
too
many
pros
and
cons
so
yeah.
If
anyone
can
give
me
more
details
on
this
will
be.
A
B
Yeah
that
that's
correct
and-
and
I
can
see
that
this
was
min-
who
is
one
of
our
engineers?
I
think
I
I'll
I'll
take
an
action
item
to
follow
up
and
report
back
on
the
pr.
A
Yeah,
please
please
do
so
so
so
what
I'm
hearing
and
what
I
read
from
this
one
is
that
the
cloud
watch
requires
to
not
do
any
aggregation
for
for
a
bunch
of
instruments.
A
B
Okay,
I'll
I'll
unders
again,
an
estimate
of
you
know
what
the
urgency
is
as
well
as
what
the
progress
is
on
our
end
for
integrating.
A
The
other
updates,
so
we
merged
the
the
initial
pr
that
added
the
gauges,
which
I
think
everyone
was
happy
with.
I
did
a
couple
of
follow-ups.
The
issue
191
is
just
a
cleanup
of
bunch
of
comments
in
the
proto,
so
there
is
a
pr
called
191,
so
people
can
review
that
and
there
is
the
most
interesting
one,
which
is
the
temporality,
which
I
would
like
to
hear
opinion.
If
anyone
has
opinion
about
this
and
how
can
we
support
a
generic
generic
strategy?
A
Slash
interval
for
for
all
the
aggregations,
so
yeah?
These
are
the
issues
that
the
two
pi
that
I
would
like
people
to
focus
on
the
next
and
most
likely
after
192
is
finished.
We
we
most
likely
have
a
working
protocol.
We
we
still
gonna
miss
the
raw
values.
If,
if
we
need
them
or
not,
it's
up
to
decision
and
right
now,
we
do
support
deltas
for
for
up
down
counters,
but
that
may
be
a
decision
in
seven
to
five
that
we
we
just
remove.
C
I
didn't
quite
I'm
just
I'm
just
looking
at
your
pr
really
quickly
now.
There's
that
comment
there
about
the
report
strategy,
and
you
say
another
approach
is
to
make
it
could
be
to
make
it
aggregation,
interval,
delta,
arbitrary
and
cumulative
you're,
saying.
A
Saw
it's
a
couple
of
systems,
it's
a
good
question.
Actually,
I
saw
a
couple
of
systems
that
support,
so
so
in
some
cases
we
want
to
report
changes
from
the
previous
report
time,
which
we
call
delta.
In
some
cases,
we
want
to
report
changes
from
the
the
last
reset
time,
which
usually
is
the
starting
of
the
process
in
some
cases.
A
A
I
saw
people
reporting,
let's
say
the
percentiles
for
the
last
10
minutes
or
for
the
last
five
minutes,
even
though
they
report
every
10
seconds
when
they
compute
that
metric
they
compute
as
what
was
my
percent
latency
percentile
in
the
past
10
minutes.
C
A
It
could
be
added,
but
but
what
I
would
would
like
to
to
to
you
and
everyone
who
is
reviewing
this
to
see
if,
if
the
right
name
for
this
is
report
strategy,
or
we
go
with
a
different
name
for
this
property,
we
know
what
it
means
we
I
try
to
explain
what
it
means
very
well,
but
maybe
the
name
we
change
it
also.
Also.
The
arbitrary
in
my
comment
was
mostly
to
to
show
people
that
we
can
extend.
I'm
not
gonna
plan
to
add
it
right
now,.
A
And
I
would
also
like
to
know
from
from
you
folks
what
do
you
think
we
miss
or
what
are
the
things
that
you
believe
are
missing.
So
these
are
the
five
items
that
I'm
focusing
right
now,
but
if,
if
there
are
other
things
that
we
you
believe
we
missed
before
having
a
working
version,
please
point
me
or
here
or
or
somewhere,
because
because
I
need
to
focus
on
on
making
sure
we
we
do.
We
we
make
a
decision
very
soon
on
this.
A
Examples
yeah
examples.
I
would
do
one
small
pr
good
good
to
know.
I
will
take
a
note
here.
I
will
just
remove
the
current
exemplars
and
wait
for
the
new
definition
to
come.
A
C
A
A
Okay,
I
think
I
think,
then
I
will
give
everyone
this
time
back.
Please
review
these
issues
that
I
posted
in
the
chat
and
the
the
the
prs,
and
I
will
focus
and
monitor
them
every
day.
So,
okay
change
any
change.
I
will.
I
will
try
to
apply
that
and
make
sure
we
we
make
progress
with
the
protocol.
Okay,.
B
Okay
sounds
good,
appreciate
it
and
and
we'll
we'll
keep
on
it.
Okay,
perfect.
Thank
you.
Thank
you.