►
From YouTube: 2020-07-23 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
Hold
up
give
everybody
another
few
minutes
to
join
in
and
if
you
have
any
issues,
just
add
them
to
the.
A
A
A
There
you
go,
you
have
so
many
issues.
You
are
going
to
talk
about
all
of
them
today
or
pr's,
rather.
B
B
B
A
I
saw
the
same
behavior
in
in
java
as
well,
so
I
don't
know
if
I
don't
know
if
there
was
some
someone
who
implemented
it
the
same
way
across
both
sides
or
if
there
was
something
in
the
spike.
C
A
B
Yeah,
but
what
sense
does
it
make
to
make
an
ore
with
a
one.
A
A
B
A
I
try
and
look
at
the
the
spec
to
see
if
there's
a
definition
for
this,
and
if
there
isn't,
I
would
raise
a
back
issue
around
it
all
right,
all
right,
so
I
don't
know
if
clayton's
on
the
call,
yet
I
don't
see
him
on
here,
but
all
right,
okay,
so
first
thing
on
the
agenda
is
I
just
wanted
to
call
out
if
there's
folks
that
are
interested
in
becoming
maintainers
on
the
python
sig
we're
we're
always
interested
in
in
more
people
that
are
that
are
keen
on
it.
A
Currently
late
in
the
night
are
kind
of
splitting
some
of
the
work
and
both
of
us
are
are
not
dedicated
full-time
to
open,
telemetry
python.
So
if
there's
other
folks
that
are
interested,
you
know
to
kind
of
work
through
becoming
an
approver
and
then
becoming
a
maintainer.
That
would
be
that'll,
be
awesome
and
if
you
don't
know
what
it
takes,
there's
a
community
guidelines
on
how
to
become
a
maintainer
and
an
equipment
project.
A
So
I
think
it's
just
a
certain
amount
of
contributions
and
reviews
to
become
an
approver
and
then,
after
that,
it's
another
amount
of
time
involved
with
the
project
to
become
a
maintainer.
D
Yeah,
I
would
probably
be
I'm
working
on
this
full
time,
not
like
specifically
python
necessarily,
but
I
think
that
would
definitely
be
a
good
goal
for
me
I
mean:
how
long
does
that
take?
Would
you
say
six
years
full
time.
A
Yeah,
I
don't
know,
I
think
I
it
took
me
about
four
months
or
so
or
maybe
three
months
to
get
to
a
maintainer.
So
it's
it's
not
that
long.
It's
just
it
just
takes
time.
If
you're
interested,
I
can
send
you
the
like
community
guidelines
on
getter.
After
this.
D
Yeah,
that
sounds
good.
I
could
probably
dig
them
up
too.
I
think
I
read
through
them
briefly,
but
okay
yeah
feel
free
to
ping
me
offline.
A
And
we
can,
we
can
chat
about
it.
Cool
thanks
laden
wanted
to
talk
about
the
next
release.
A
I
don't
know
what
he
wanted
to
say
about
it,
so
we're
just
gonna
skip
to
the
next
and
the
next
agenda
item
until
he
gets
here.
If
he
does,
I
just
wanted
to
follow
up
on
the
ga
spec
reviews
from
last
week,
so
laden-
and
I
did
some
triaging
on
the
open
issues
on
the
project
and
we
also
created
issues
for
the
for
reviewing
the
spec
and
our
alignment
with
us
back
in
these
different
components
and
so
for
the
ones
that
him
and
I
were
assigned
to
we've
already
sent
ourselves.
D
Yeah,
that's
totally
fine.
I
didn't
have
a
chance
to
take
a
look
yet,
but
I'm
definitely
planning
to
do
it.
A
Cool
hey
diego:
do
you
want
to
take
on
the
sampling
one
since
you're
already
looking
at
sampling
things.
A
All
right,
if
nobody
has
another
issue
or
another
topic,
we
can
dive
into
the
top
into
the
pr's
here.
D
Yeah,
I
don't
know
if
leighton
is
going
to
show
up,
but
I
think
you're
just
wondering
when
the
next
release
is
going
to
be.
That's
all.
A
D
Yeah
I
mean
you
probably
know,
since
you
said
that
you
were
working
on
the
the
an
exporter
yourself
you're
like
maintaining
one,
but
I
think
it's
been
like
almost
a
month
or
maybe
over
a
month
since
the
last
one.
So
we
can't,
I
mean
we
could
release
like
bug
fixes.
But
I
guess
maybe
just
is
there
like
a
really
scheduled
plan
so
that
we
can
plan
our
releases
as
well.
A
So
we
have
been
kind
of
going
on
a
by
bi-weekly
cadence
on
the
window.
A
But
I
don't
know
exactly
what
that's
going
to
be
so
yeah.
I
mean,
I
guess.
If
we,
if
we
keep
to
like
our
two-week
cycle
it,
it
seems
to
be
pretty
straightforward
and
there's
not
that
much
work
to
do
so.
A
C
I
just
I
want
to
kind
of
just
check
in
when
we
say
next
release.
It
just
takes
everything.
That's
in
master,
correct.
A
A
B
E
I
have
one
approval
and
late
and
reviewed,
and
I
addressed
his
comments,
so
I
only
use
the
second
approval
for
this
and
I
appreciate
getting
it
in
sooner
rather
than
later,
so
I
don't
have
that
much
time
left.
A
A
A
Just
gonna
look
at
the
notes
and
wonder
why
you
missed
this
meeting,
but
all
right
instrument,
instrumentation.
A
C
Yeah,
I
added
that
I
think
that
was
one
of
the
pr
that
we
were
blocked
on,
so
we
just
this
one.
Okay,
let's
go
out
so
here
I
got
my
answer.
C
C
And
so
this
was
also
added
by
me,
actually
kevin
from
one
of
my
team.
Other
teams
added
this.
I
don't
have
much
context
on
this
and
I
was
hoping
that
he
would
be
here.
I
guess
he's
not
in
the
call,
so
I
don't
know
if
somebody
can
update
some
on
this
issue.
That
would
be
great.
I
I
just
don't
have
much
context
on.
A
C
Yeah,
I
don't
think
we're
running
collector,
but
you
know
again
I
have
to
confirm
because
he
was
the
one
working
on
this
project,
but
my
from
what
I
know
I
don't
think
we're
a
running
collector.
Yet,
okay.
F
This
by
the
way
hi
to
here
here,
the
this
is
something
that
we
had
talked
about.
I
talked
about
with
sugar
a
few
weeks
ago
and
I
I'd
be
very
interested
to
see
what
others
think
about
how
to
support
something
like
this.
You
know,
I
think
we
might
have
had
a
chat
about
this
in
the
sig
meeting
as
well,
but
just
like
what
there's
nothing
in
the
api
or
sdk.
F
That
really
supports
this,
but-
and
I
guess
it's
also
connected
to
the
chat
last
week
about
the
you
know-
pii
issues
with
db
statements,
but
I
suppose
in
this
case
the
you
know,
what's
what's
whether
there's
something
that
folks
could
do
or
whether
there's
some
development
work,
that
we
think
we
might
want
to
add,
at
least
in
python,
to
support
it.
F
Yeah,
which
is
yeah,
which
is
just
like
you
know,
when
I
looked
at
it,
a
span
processor
made
some
sense
right,
a
span
processor
that
would,
in
addition,
look
at
spans
before
they're
actually
exported
it's
not
as
I
don't
know.
If
that's
the
right
approach
here,
though,
mainly
because
spam
processors
aren't
called
in
succession
necessarily
they're,
it
is
especially
with
the
what
is
the
multi-span
processor
or
concurrent
span
processors,
so
yeah.
F
So
is
there
any
really
like
filtering
mechanism
that
we
could
add
to
support
something
like
this
or
is
it
something
like?
If
someone
wants
to
do
this,
they
really
have
to
do
this
at
exporter
level.
A
Right
yeah,
so
I
know
the
I
know
the
collector
has
been
implementing
some
amount
of
like
provide
the
ability
to
replace
values
inside
like
attributes
and
and
stuff
like
that,
for,
like,
I
guess,
for
handling
some
amount
of
scrubbing
of
data.
A
I
don't
know
if
the
solution
of
the
collector
is
like
the
only
solution
that
would
be
applicable
in
this
case.
Clearly
not
everybody's
going
to
be
using
a
collector.
So
you
know
it
would
make
sense
for
different
implementations
to
support
something
like
this,
but.
C
C
In
on
yeah,
people
should
be
able
to
just
get
in
there's:
okay
yeah
I'll.
Just
maybe
he's
got
the
wrong
meeting
or
something
sorry
about
that.
A
First
yeah:
it's
a
maybe
the
takeaway
from
this
because
I
think
to
hear
you
have
you
had
a
chance
to
look
through
any
of
the
other,
either
implementations
or
respect
to
see
if
anything
like
this
is
talked
about.
F
Yeah
I
looked
at
it
before
and
I
don't
I
didn't
see
it
I'll
look
at.
Maybe
if
you
think
there's
another
client
library
that
might
have
it
but
like
javascript,
I
imagine,
might
have
enough,
but
I
think
yeah,
if
it's
happening
in
the
collector.
That
makes
sense,
I
think,
having
a
solution,
at
least
in
this
case.
For
you
know,
in
the
case
of
the
data
exporter,
I
think
I'm
gonna
have
to
think
about
how
to
do
that.
F
I
mean
it
might
be
just
like
adding
adding
an
additional
filtering
mechanism
to
the
exporter
so
that
it's
not
really
just
a
hacky.
It's
just
like
okay.
This
exporter
just
supports
additional
filtering,
though
this
is
not
something
that's
particular
to
one
any
one:
exporter
I'll
check
with
the
yeah
I'll
check
into
the
collector
and
see
what
that
looks
like.
D
Hey
so
this
is
kind
of
similar
to
the
spam
processor
idea,
but
I
mean
it's
also
kind
of
hacky,
but
you
could
make
like
a
an
exporter
that
wraps
your
like
datadog
or
whatever
exporter,
and
then
that
one's
export
would
filter
out
the
spans
before
sending
them
on
to
to
like
the
export
that
it
wraps.
Would
that
work,
or
I
mean
it's
not
like
it's
not
really
clean,
but
would
that
do
the
job.
F
Yeah,
I
mean
yes,
so
I
think
you
think
you
had
a
similar
idea
where,
like
yeah,
the
exporter
calls
a
processor
yeah.
Something
has
a
pre-processing
that
would
happen
in
the
exporter.
F
I
just
wonder
if,
like
that,
something
like
a
filtering
out
pii,
I
imagine
someone
would
want
that
sooner
rather
than
later,
but.
F
Processor
is
a
span
processor,
the
right
way
to
think
of
this
right
and-
and
even
I
guess,
questions
like
what
is
a
span
processor
right
like
it
seems
like
a
span
processor,
the
two
that
are
there
are
like
ones
that
do
that
sort
of
some
form
of
simple
or
batching
right,
but
is
there
a
generalization
of
a
span
processor
that,
like
could
do
things
with
spans
right
and
mutate
them
alter
them
change
that
you
know
change
such
span
processors
so
that,
like
they're,
applied
in
sequence
or
in
some
some
fashion.
F
E
A
E
A
Definitely
in
the
in
the
collector,
you
can
chain
processors,
oh
interesting,
so
you
can
change
process
yeah.
So
I
mean
I
would
suspect
that
that's
something
that
you
should
be
able
to
do
so
you
could
have
some
kind
of
like
filter
span,
processor
that
passes
into
the
batch
span
processor
or
something
like
that.
E
A
Because
you
can
definitely
add
multiple
processors
to
like
a
trace,
tracer
provider.
But
I
don't
know
what
the
implementation
is.
As
far
as
how
the
like
the.
F
Yeah,
the
flow
of
the
data
goes,
the
simple
span,
processor.
What
it
does
is
it
applies
it
in
sequence,
so
there's
just
a
list
and
the
same
span
will
be
fed
on
so
like
you
can
meet,
you
can
have
position
zero
mutate,
the
span
so
that
the
span
processor
at
the
end,
let's
say
that's
the
one
that
actually
like
exports
it
right.
That's
the
one
that's
exported
will
be
mutated,
but
we
also
have
this
multi-concurrent
one
where
a
span
is
sent
to
all
processors
at
once,
and
so
like
I
don't
know.
F
If
that's
that's
that
doesn't
seem
like
it's
gonna,
be
the
right
place
to
do
something,
but
chaining
them
would
be
like
the
most
obvious
place
to
do
it
or
like
having
a
pre
processor
of
some
form.
Again,
like
I,
you
know
I'm
you
know
I
don't
really
have
enough
context
but
like
what
is
a
spam,
how
are
people
using
spam
processor
or
how
do
people
imagine
people
folks
using
spam
processors
in
their
applications.
G
So
this
is
kevin.
I
opened
the
issue,
so
we
were
working
with
yusuke.
I
think
he's
one
of
the
maintainers
or
developers
on
open,
telemetry
python
and
the
spam
processor
was
his
first
way
of
trying
to
bring
this
functionality
from
dd
trace
over
to
using
open
telemetry
instead.
So
it's
definitely
our.
I
think
he'll
just
first
go
at
it
and
from
what
we
saw
well.
A
No,
I
think,
we're
I
think
it's
pretty
clear
what
the
what
the
issue
is.
I
think
we're
we're
just
trying
to
figure
out
what
what
the
best
way
to
go
about
solving
it
is.
G
And
so
for
us
right
now,
since
the
default
scan
processor
is
the
synchronous
one,
we
have
to
make
sure
that
our
filtering
is
occurring
before
the
datadog
expert
spam
processor
is
called
since
its
on
end
function
is
going
ahead
and
actually
exporting
it
to
datadogs.
So
looked
at
the
concurrent
one,
it
would
be
too
since
they're
all
async.
I
guess
it
would
be
too
late
to
actually
alter
and
remove
the
pii
or
scrub
the
pii
in
time.
So
synchronous
one
does
seem
like
the
right
approach
here.
F
G
F
Yeah,
so
I
mean
if
it
if
it
makes
sense,
I
can
check
in
with
a
collect
check
into
the
collector
alex's
suggestion
about
the
specs.
What
he's
looking
at
right
here
and
then
yeah
we
can
just
take
this
into
the
issue,
chat
the
issue
that
they've
created
and
hopefully
make
some
progress
on
that
awesome.
A
Yeah,
I
think,
based
on
what
the
spec
is
showing
here,
it
does
seem
like
that
would
be
the
only
place
where
we,
or
at
least
the
intended
place
of
where
filtering
will
be
done.
So
I
mean
it's
probably
interesting
to
see
what
the
collector
is
doing
as
far
as
how
it's
applying
any
kind
of
processing
it
does
and
see.
If
we
can
do
something
like
that
in
in
python.
A
A
A
All
right,
I
didn't
really
have
any
other
issues.
I
don't
know
if
some
folks
have
any
open
pr's
that
I
want
to
talk.