►
From YouTube: 2020-10-30 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
B
C
B
A
B
Yeah
yeah,
no,
it's
it's.
I
think
it's
you
know
I
think
earlier
this
year
I
was
pretty
optimistic
with
my
timetable,
but
at
the
same
time
like
I
don't
know,
I'm
really
happy
with
the
work
we've.
B
A
A
D
A
We
should
get
rolling
here
if
we
have
enough
here
to
to
make
choices.
Yeah
all
right
this,
let's
I'm
just
going
to
work
upwards
on
this
list,.
C
A
F
G
B
G
G
B
Yeah,
but
that
would
technically
make
anything
we
do
later.
A
non-breaking
change,
yeah,
okay,
put
put
in
there
like
we
enter.
If
you
consider
already
writing
just
and
say
like
setting
making
port
number
required
parameter,
would
allow
us
then
to
set
it
by
default
later
to
any
port.
We
choose.
B
I
think
I
think
we
either
do
that
or
we
just
say
no
we're
sticking
with
that
port
number
too
bad.
G
B
Yeah,
what
what
port
number
we're
using
again?
Can
you
scroll
up
five,
five,
something:
five,
five,
six,
eight,
I'm
on
the
ina
site.
B
G
B
A
G
E
F
Yeah
riley
is
this:
how
do
you
see
this
getting
manifested
like
like?
Does
this
pop
up
somewhere
for
you
like.
F
Yeah
and
go,
we
do
the
same
thing
like
we
just
kind
of
made.
I
don't
know
whoever
made
the
pr
got
to
choose
that
one
yeah
yeah,
I
just
didn't
know
like
I
downstream.
Does
this
actually
show
up?
I
guess
is
my
question.
D
D
And
my
impression
is
that
things
like
we're
not
making
changes
like
this
one,
which
feels
like
literal
changes
they
could
be
allowed
for
gi
not
required,
and
my
worry
is
that
I
feel
that
I
feel
personally
that,
as
long
as
we
have
required
for
gi
issues,
we
won't
be
able
to
freeze
this
back.
A
Related
to
that
related
to
what
carlos
is
pointing
out,
we
can
address
it
after
we
can
dive
more
into
it
after
we
do
our
triage,
but
I
do
have
a
question
about
our
velocity
like
p1
is
coming
in
fast
like
oh,
you
have
this
small
change
like
if
it's
quick
and
we
can
get
it
in.
We
can
finally
conclude
our
freeze.
We
got
two
outstanding,
but
we
can
talk
about
that
after
this,
but
I.
G
So
what
is
our
strategy
before
before
pressing
the
buttons
here?
What
is
our
strategy
related
to
allow
for
ga?
So
so
there
is
going
to
be
a
freeze.
Okay,
there
is
going
to
be
a
freeze,
let's
say
next
week.
We
should
have
the
freeze
for
for
trades,
but
there
are
a
bunch
of
things
that
we
marked
as
allow
for
ga.
Are
we
going
to
call
ga
without
them?
I
mean
yeah?
Yes,
yes,
this
can
happen
without
them
for
sure,
but.
G
I
think
we
applied
different
rules.
We
we
said
we
said
that
if
it's
not
a
breaking
chain,
so
it's
just
an
addition
to
whatever
we
have
right
now
we
can
still
freeze
the
spec
because
we
can
add
a
bit
more
things,
but
now
now
now,
when
you
say
like
dcga,
really
is
like
super
ready
and
whatever
you
call
gui
some
of
the
things,
I
may
have
decided
that
I
want
them,
but
I
I
don't
think
they
are
critical
enough
for
these
feeds
for
for
the
release
candidate,
but
we
can
still
polish
things
so
anyway.
B
A
C
A
Candidates,
we
wouldn't
look
too
much
at
the
p3s
yeah,
but
if
we
do
identify
something
after
the
freeze
we're
like
hey,
we
really
got
to
get
this
in.
We
promote
it
to
a
p1
in
order
to
yeah.
G
G
A
G
G
Yeah,
it's
a
hypebeast
makes
sense.
Okay,
so
that's
how
I
would
describe
it
and
I
don't
know
how
we're
gonna
do.
Probably
the
right
thing
to
do
is:
okay,
let's
stick
with
p1
require
4g
for
things
that
are
breaking
and
we
really
need
to
have
them
right
now
and
then,
maybe
maybe
the
week
after
we
we
freeze
that
we
go
over
all
allow
for
ga
and
we
do
a
retriage
of
p2s
some
of
them.
We
mark
as
p1,
allow
for
ga,
and
we
say
okay
for
the
final
g.
We
want
all
the
p
ones.
A
Yeah,
if,
if
the
change
comes
in
as
additive,
you
know
it's
like
or
it
has
like
the
optional
keyword
or
something
like
that
in
there
I
mean
the
change
of
the
spec
makes
it
optional
for
the
implementation
of
the
language,
specs
and
they're
super
easy
to
slip
in
those
could
be
promoted
to
p1.
I
would
see
my
mind
if
it's
a
super
breaking
breaking
change,
then
we'd
have
to
like
really
go.
We
missed
this
one
and
we
have
to
do
an
rc2
or
something
yes,
because
we
hey
guys
we
had
to
change
the
api.
A
B
H
B
Just
following
up
from
the
first
issue
that
we
looked
at
the
ina
iana
says
it
takes
about
three
weeks
to
get
a
port
registered
and
we
appear
to
meet
the
requirements.
So
I'm
gonna
I'll
just
apply
for
this
today.
I'll
see
tigran
because
he's
pretty
involved
in
the
protocol
and
see
if
we
can
get
something
quickly,
yep
sweet
all
right
change
this
to.
G
G
E
So
I
have
a
question
I
wonder
like
if
we
could
give
some
flexibility
to
individual
language,
for
example
like
some
languages,
like
c
plus
plus,
if
you
like,
started
by
returning
an
integer
and
literally
change
that
to
like
just
for
better
like
static
attack.
That's
not
considered.
E
G
G
Yes,
if
I
heard
correctly
the
initial
part,
he
was
mentioning
that
there
there
is
different
language,
the
stuff
which
I
agree,
but
legal
nutrition.
G
Because
we
want
people
to
whatever
extend
the
the
the
api
so
you're,
not,
I
don't
think
I
don't
think
this
is
related
to
to
what
right
you
mentioned
rightly
mention
some
compatibility
like
when
you
do
a
change
in
a
language
by
changing
a
return
type
from
in
to
a
noun
or
something
like
that.
G
G
Should
we
do
this,
so
I
think
I
think
we
should
just
clarify
that
the
api
does
not
define
the
return
types
intentionally,
because
we
want
to
allow
alternative
implementation
that
will,
for
example,
record
the
timestamp
every
time
when
you
put
an
attribute
or
other
crazy
things
that
they
want
to
do
so.
E
Yeah,
I
I
think
look
at
the
spec.
It
just
gave
the
semantic
like
what
what
kind
of
the
feature
you
should
provide,
but
it
it's
never
trying
to
restrict
you
or
like
to
follow
the
strict
like
function,
name
or
strict
type.
So
so,
instead
of
tagging
this,
I
think
we
can
probably
just
respond
and
close
this
issue.
G
G
G
H
A
A
G
G
This
is,
this
is
important
for
our
yaga
exporter
and
stuff.
I
would
say
it's
allowed
for
ga
p2,
because
we
need
the
two
to
figure
to
support
jager
based
on
our
it's
an.
G
H
G
A
Perfect,
thank
you,
and
there
was
one
okay.
We've
got,
the
p1p1p1p1
needs
an
assignee.
A
We've
got
three
that
are
not
metrics
related.
This
protocol
trace
trace
so
just
want
to
highlight
two
open
pr's
and
I'm
going
to
bring
the
topic
up
again
on
the
maintainers
meeting
on
how
we
can
finalize
our
trace,
spec
freeze
and
I'm
willing
to
help
with,
like
the
tagging,
the
change
log
on
the
spec
sig
repo.
If
that's
helpful,
in
order
to
be
able
to
include
those
things.
A
14
15
p1s
and
we've
been
working
diligently
to
assign
them
all.
So
that's
kind
of
the
momentum
that
we're
trying
to
get
towards.
G
E
G
F
Do
you
know
morgan
out
of
the
three
weeks
just
for
context?
Do
you
do
they
kind
of
like
give
you
an
idea
what
port
it's
gonna
be
and
then
they
finalize
it
over
the
three
weeks
or
it
takes
three
weeks
to
get
a
number
essentially.
B
F
C
F
G
Yeah
good
point,
because
you
you
put
a
default
there,
unless
you
required
users
to
tell
you
the
port.
B
H
I'm
thinking
we
bundle
up
the
collector
as
a
docker
container
that
docker
container
we
can
kind
of
define
whatever
port
we
want,
because
we
have
a
lot
of
control
over
that
container
yep.
We
can
maybe
move
that
default
out
of
the
collector
code
base
itself
and
into
the
definition
of
the
docker
container.
B
F
B
B
G
B
E
I
am
very
supportive
of
that,
so
just
one
question,
probably
not
directly
right
here
here,
so
I've
seen
like
the
environment
variable
related
to
the
sampling.
Well
like
doing
some
experimental
work
on
the
logging
part.
I
I
think
the
the
server
interface
currently
is
very
specific
to
trace
and
I
like,
if
you
look
at
the
current
sampler,
it
takes
all
the
sampling
parameters
from
the
trace
part.
Imagine
like
you
will
have
the
center
for
logs
and
when
we
design
the
environment
variable,
we
got
to
consider
they
might
be
taking
different
samplers.
E
G
Just
just
just
say
that
we
should
put
trace
inside
the
name
for
for
making
sure
that
if
we
ever
add
another
sampler
in
other
package,
don't
name
logging
or
anything.
Just
say
that
if
we
have
a
sampler
in
other
packages,
even
metrics,
for
example,
we
can
have
sampling
to
not
shoot
ourselves
in
the
foot.
So.