►
From YouTube: 2020-08-25 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
B
C
B
Yeah
well,
what
we've
done
is:
we've
said
that
the
metrics-
sorry
not
metrics,
the
tracing
spec
needs
to
be
finished
next
week
or
the
governance,
not
governance.
The
technical
committee
will
just
close
out
the
final
issues
themselves,
like
just
very
rapidly
like
basically
no
more,
no
more
discussions
or
debates
after
next
week.
B
So
after
that,
then
the
first
wave
of
sdks
can
have
their
tracing
release
candidates
pretty
quickly
once
they're
done,
then
we'll
do
the
same.
We're
going
to
do
the
same
next
for
all
of
the
other
apis
that
are
related
to
tracing
like
context,
propagation
and
others,
and
we'll
give
that,
like
a
two
week
deadline
or
something
we'll
debate
it
on
monday
and
then
finally,
we'll
do
it
for
matrix.
B
We
will,
on
monday
the
the
tracing
one
has
the
date
set,
it's
the
first,
the
first
one
we
did
this,
for.
I
think
people
are
pretty
happy
with
it,
but
I
I
want
to
start.
We
said
the
first
week
of
september
was
the
deadline
and
I'd
like
to
sort
of
solve
it
for
the
others.
On
the
first
week
of
september,.
B
B
B
I
don't
think
so
I
I
I
I
can't.
I
think
sampling
is
part
of
the
tracing
api,
so
probably
not
yeah,
at
least
not
new,
like
specifications
or
anything.
Okay,.
B
But
next
monday
and
tuesday
is
when
we're
going
to
go
over
the
the
any
remaining
work
if
there
is
any
on
tracing
okay,
to
be
honest
after
setting
the
deadline,
I
think
things
have
really
sped
up
so.
C
B
A
B
After
this
week,
okay
or
after
next
week,
yeah
and
we'll
certainly
be
adding
more
like
the
ga
is
going
to
be
like
the
minimum
set
of
things
we
need.
There
will
be
apis
that
ship,
post,
ga,
okay,
right
and
and
there'll
be
conventions
that
get
decided
on,
or
at
least
more
concretely
decided,
post.
Ga.
B
A
Seen
while
you're
bringing
that
up,
I
saw
one
just
looking
through
the
this
morning's
agenda.
One
of
the
things
was
initial
requirements
for
integration,
testing,
which
I
haven't
really
read
through
yet.
B
B
Oh
no,
that
was
after
I
joined
okay,
looks
like
there's
a
github
issue
for
by
riley.
B
B
B
C
B
C
C
C
C
C
C
B
B
C
C
B
There
was
not
okay
very
few.
These
had
a
discussion.
The
meeting
ended
at
about
half
40,
so
most
of
the
discussion
was
andrew's,
two
topics
which
were
not
time
box,
but
that's
that's
fine,
which
was
just
the
the
overview
of
what
still
needs
to
be.
A
C
B
Yeah,
so
not
all
of
these
were
discussed.
This
is
basically
just
like
andrew
went
and
pasted
the
got
it
yeah.
Okay,
you
guys
talked
about
the
multiple
tracks
that
we
provided
discussion.
A
B
We
did
talk
about
that
yeah,
so
the
discussion
for
that
was
going
to
continue
on
github.
There
was
still
some
disagreement
over
that
one
over
whether
it
was
a
necessary
scenario
which
I
think
most
people
believe
it
is
and
whether
we
should
support
it
for
the
rc
and
ga
or
whether
it
can
be
added
later
as
a
non-breaking
change.
B
I
think
it's
necessary,
but
for
if
it
doesn't
shift
for
ga,
that's
okay,
it's
not
the
end
of
the
world,
and
then
there
was
further
discussions
that
are
going
to
take
place
on
github
and
how
to
do
it.
There
was
like
whether
you
registered
a
separate
tracer
or
whether
there'd
be
like
a,
I
think,
like
a
mock
sort
of
tracer,
that
you
would
then
register
on
top
of
some
wacky
thing
like
that.
B
B
Yeah
yeah
here
we
go.
Here's
the
christian
head
weighed
in
this
was
discussion
this
morning
like
he
was
saying,
exporters
must
not
be
attached
to
more
than
one
span
processor
unless
it
explicitly
states
that
it
supports
it
and
span
process
cannot
be
attached
to
more
than
one
tracing
provider
unless
it
specifically
explicitly
supports
it.
Yeah.
I
guess
that's
the
challenge
like
with
open
census.
I
don't
think
we
had
any.
I
don't
think
we
have
an
equivalent
of
spam
processor
back
then,
like
the
exporters.
C
A
B
There's
some
basic
use
cases
like
there's
some
sort
of
really
big
enterprises,
like
I
know
from
like
talking
with
them
directly
and
talking
to
microsoft.
People
talking
from
like
walmart
in
particular,
has
like
imagine
like
20,
well,
hundreds
of
teams,
but
each
team
uses
some
sort
of
subset
of
the
same
sort
of
master
tool
set,
many
of
which
overlap,
and
so
you
may
have
like
a
team
that
uses
jager
for
tracing
and
one
that
uses
dynatrace
for
tracing
and
one
that
uses
zipkin
for
tracing,
and
so
they
have,
for.
B
I
think
much
of
walmart
standardized
their
instrumentation
at
least,
but
they
would
need
to
be
able
to
send
this
data
to
multiple
backends
at
the
same
time,
so
that
multiple
teams
can
use
different
tools.