►
From YouTube: 2019-10-08 Spec SIG
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
C
C
C
C
C
C
D
D
D
E
E
A
D
A
The
way
you
know
to
not
because
because
people
are
worried
that
you
will
forget
to
do
a
separate
class
and
stuff
probably
is
better
to
just
simply
create
issues.
If
there
are
something
that
you
you
decide
that
you
will
do
in
a
separate
er,
just
create
a
separate
issue.
Link
that
issue
to
the
comment
and
say:
hey,
we'll
address
this
here
and
just
like
resolve
that
comment,
so
that
you
make
sure
that
you
have
always
all
the
comments.
Closed.
E
We
there
were
a
few
issues
open
some
buy
me,
some
better
reviews
themselves
and
or
linked
in
the
comments.
I
think
yeah,
maybe
I
should
trust
mark
them
as
resolve
the
Tres
yeah
yeah.
A
E
D
E
D
E
G
A
A
E
G
E
We're
needed
to
set
to
two
something's
to
the
Spain
context,
set
and
really
supported,
although
in
person
you
can
just
monkey
patch
like
you
want,
we
probably
want
use
this
approach.
So
it's
it's
possible
that
we
might
not
even
need
that,
although
it's
yeah,
although
it
will
be
good
to
not
have
to
use
the
string
based
mapping
for
everything.
D
A
Think
a
lot
languages
have
have
the
concept
of
one
even
see
past
class.
In
couple
of
them
there
is
a
question
of
also
it
will
allow
people
to
extend
that
class
and
being
sort
of
hot
and
I
want
to
see
some
real
use
cases
I
mean
theoretically,
I
can
see
use
cases,
but
that's
why
I
was
asking.
Is
this
a
real
use
case
where,
where
dynaTrace
really
needs
this-
and
it
seems
seems
to
me
from
the
answer
that
there
is
some
kind
of
real
beauties
here.
E
A
Breaking
yeah,
we
we
can
do
even
breaking
changes
until
one
zero,
so
we
said
there
is
no
problem.
If
there
is
a,
we
discover
that
this
is
a
real
problem.
You
do
it.
We
just
need
to
to
understand
if
this
is
a
real
problem,
how
do
L
read
out
that?
But
in
general
I
like
to
also
know
I
mean
we
can
come
up
with
theoretical
reasons
for
a
lot
of
things,
but
I
want
to
see
some
some
use.
A
A
A
If
your
argument
is
just
about
allowing
people
to
extend
these
plans,
I
don't
think
it's
a
very
hard
argument
to
provide
like
if
the
argument
is
to
make
it
everything
abstract
you
will
have
I
can
have
argument
against
that.
But
if
you,
if
you
just
want
to
extend
that,
then
it's
probably
fine
but
anyway
I
think
we
should
read
the
author
before
commenting
yeah.
H
Well,
by
the
way,
every
job
and
about
making
it
as
an
opaque
object,
you
know,
like
you,
don't
expose
anything
I
would
became
that
we
did
in
a
bit
racing
as
we
mentioned
so
yeah,
and
we
used
to
expose
the
ideas
of
strings
like
both
traits
ID
and
Spinelli,
so
yeah
anyway.
Let's
comment
more
down
the
topic:
what
yeah.
D
Great
great,
please
have
a
look
at
that.
Moving
on
what
looks
to
be
the
kind
of
final
major
piece
in
the
major
metrics
API.
There's
a
label
set
otep
still
outstanding.
This
is
related
to
the
large
change
in
metrics,
that's
currently
being
proposed
in
the
spec
Rico.
This
is
the
one
portion
that
hasn't
made
it
through
the
o
tab
phase.
Yet
I,
don't
know
I,
think
Bogdan.
A
D
A
D
Moving
on
sounds
like
there's
a
meeting
already
I,
don't
know
if
there's
one
on
one.
If
a
meeting
does
end
up
on
the
calendar,
let
me
know
I'm
happy
to
bring
more
people
in
moving
on
the
implementing
named
tracer
metrics
this.
This
is
made
it
through
the
spec.
This
is
now
in
the
spec,
but
I'm
not
sure
if
this
has
been
implemented
anywhere
yet
so
I
just
wanted
to
raise
awareness
of
this
named
tracers
named
metric
feature.
That
was
just
added.
Is
your
Sigma
attained
er?
Please
have
a
look
at
this.
D
G
I
followed
our
guidelines
or
relative
guidelines
and
I
created
issues
for
all
the
language
repost
there.
So
I
think
they
most
people
should
be
aware
now
most
maintain
us
that
they
should
implemented
and
I
did
a
implementation
for
the
for
dotnet.
That
being
said,
I
only
did
a
name
traces
for
named
metrics.
G
Specification,
I
guess
so
anybody
who
wants
to
start
implementing
another
language
read
book
and
maybe
look
at
its
top
net,
how
its
implemented
there
I
also
got
a
question
already
from
Ruby
from
Ruby
a
repository
against
it
was
yeah
and
they
were
asking
about
an
example
or
how
they
should
implement.
This
specification
change
and
I.
Think
Java
I
was
already
started
by
someone
from.
D
D
D
There
is
a
pile
of
us
who
felt
like
there
were
a
lot
of
edge
cases
related
to
having
some
kind
of
default.
Behavior
enabled
and
we'd
prefer
just
to
start
with
a
no
op,
since
it
doesn't
seem
like
we
have
time
to
actually
map
out
what
all
these
edge
cases
are
and
what
we
want
to
do
about
them.
Having
it
just
turned
off
for
the
time
being
seems
like
the
safest
thing.
A
D
G
I
There's
a
us
our
connection
with
global
initialization
questions
like
if
you
start
operating
on
this
thing
before
there's
an
SDK
installed
and
then
and
an
on
no
op
gets
installed.
What
do
you
do?
There's
sustaining
RFC
or
Oates
at
number
five
that
has
been
stalled
for
two
months.
I
think
we
should
talk
about
that
sometime
at
some
point.
Okay,.
D
Let
me
see
if
I
can
put
a
link
yeah
and
again
we
don't
have
Sergei
on
occult.
We
don't
need
to
hash
it
out
right
here,
but
it
would
have
raised
it
where
around
this
issue,
because
it
seems
like
some
P
havior,
it's
not
the
most
important
behavior
4.2.
If
it
was
uneven,
when
we
did
release
it's
fine,
but
it
seems
like
something
we
should.
We
should
sort
out.
D
D
J
Far
away
yes,
so
I've
been
told
that
we
need
to
move
the
exporters
all
have
the
core
SDK
repo.
So
I
want
to
do
that
as
far
as
possible.
I
want
to
know
some
guidance
from
you
guys
should
we
move
that
totally
outside,
like
github,
slash,
Microsoft
or
still
like?
We
put
that
under,
like
github,
slash,
open,
telemetry,
slash
lender
or
something.
B
Last
meeting
was
that
open
source
exporters
can
live
inside
they
open
telemetry.
We
feel,
as
almost
people
are
okay
with
accepting
that
open,
telemetry
contributors
or
the
primary
were
yours
for
it,
but
I
think
certainly
for
proprietary
explorers.
Those
should
live
in
the
organization,
maintaining
the
proprietary
code
proprietary
or
like
open
source
that
only
works
with
one
specific
service,
etc.
Okay,.
B
J
B
Do
not
create
an
under
open,
telemetry
fright.
The
reason
is
that
for
one
there's,
a
bunch
of
stuff
I
think
in
our
contribution
guidelines
that
say
that
if
you're
a
owner
of
a
open,
telemetry
repo
that
you're
authorized
to
do
certain
things
for
the
community
right,
I,
think
that
would
be
weird
to
have
vendors
as
owners
in
I
see
that.
J
K
A
The
way
we
are
doing
that
before
one
zero,
so
if,
if
you
really
want
to
test
it
now,
because
we
do
not
publish
artifacts
and
stuff,
you
may
put
it
there,
but
when
we
reach
one
zero,
we'll
kick
everyone
out,
then
the
deadline
to
to
move
everyone
out
is
one
point:
zero
before
one
point:
zero,
it's
up
to
the
maintainer
of
the
specific
language
to
decide
if
they
want
to
keep
it
there
for
testing
and
for
letting
people
try.
But
after
one
point
zero,
nothing
will
be
there.
Colin.
J
A
A
A
D
D
My
long
someone
was
asking
about
my
context
prop
PR
in
general.
That's
in
point
three,
so
I've
been
focusing
on
trying
to
get
all
these
point.
Two
issues
closed
up
and
over
the
finish
line.
So
sorry
I
haven't
made
another
push
against
that.
But
did
someone
have
a
specific
question
about
context?
Propagation,
great.
B
D
Are
we've,
we've
tried
to
go
back
through
and
prop
Sergey,
and
and
put
these
things
in
milestones.
Oh
I
see
yeah
that
pulls
about
a
milestone
over
here
interesting
we've
mostly
been
putting
stuff
into
milestones
on
this
spec
side,
but
that's
a
good
point.
Maybe
stuff
should
get
labeled
in
Oh
tips.
If
we
know
that
there's
a
point
release
of
this
spec
that
this
is
related
to
we
could.
We
can
make
a
note
here.
That
seems
like
a
great
thank
you.
C
Just
I
just
added
one
more
about
the
spec
milestones,
so
we've
got
this
massive
stock
in
the
specs
repo,
but
we
don't
really
distinguish
between
these
specs
milestones
and
the
milestones
in
the
language,
clients
and
their
releases,
so
I
mean
just
looking
at
the
dates
for
the
Alpha
v2.
It
looks
like
we.
We
had
published
a
date
that
was
next
Monday
for
the
VT
release
and
we
still
have
a
lot
of
stuff
in
the
v2.
D
D
Mean
it
would
be
nice
to
say
everyone,
please
get
it
done
by
a
certain
deadline,
but
everyone's
like
a
pretty
vague
concept.
It's
true
that
when
we
say
everyone
up
to
this
point,
there's
like
five
language
isn't
particular
that
we've
been
trying
to
like
make
sure
get
over
the
finish
line,
but
it's
definitely
not
everyone.
We're
not
saying
Ruby
and
PHP
and
Erlang
need
to
have
something
done
by
a
certain
date.
D
C
D
C
If
you're,
just
talking
about
Java
and
I,
don't
think
you
can
get
a
meaningful
release
out,
you
know
say
next
next
Monday
for
the
the
Alpha
b2,
so
I
think.
If,
if
we
want
to
coordinate
these,
we
probably
need
to
come
up
with
a
like
concrete
set
of
things.
I
want
the
clients
to
do
and
then
another
date
to
release
the
next
version.
I
also,
you
know
I,
don't
know
for
sure
that
we
need
a
release
for
each
of
these
alpha
milestones
in
the
spec
I.
D
Feel
like
it
it's
more,
the
most
important
thing
is
just
to
have
clarity,
I
think
it's
fine
for
us
to
make
a
release
announcement
where
you
say,
hey
point
to
the
spec
is
released
and
then
the
next
question
everyone's
gonna
have
is
like.
Well,
when
is
it
available
in
my
language
as
long
as
it's
it's
like
a
clear
path
to
go
to
the
repo
and
look
here,
and
you
can
see
like
this
group's
expectation,
I,
think
that
would
be
fine
and
that,
rather
than
having
some
like
cross
language,
coordinated
release,
strategy,
okay,.