►
From YouTube: 2019-09-26 Go 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
B
A
B
Stephen
I'm
not
sure
if
you've
actually
met
apologies,
we
met
before
but
nice
to
meet
you
otherwise
yeah.
C
B
E
E
B
E
E
Bt
propagator
I
need
to
reef
factor
based
on
the
commit
that
has
been
done
for
the
propagator
interface
I'll,
get
to
that
either
today
or
tomorrow.
So
we're
gonna
leave
that,
as
is
let's
look
at
the
issues,
let
me
sort
it
out
by
the
alpha
release.
So
yesterday,
night
I
put
the
Alpha
milestone
for
the
issues
that
we
need
for
September
30th
and
let
me
first
look
at
the
trace
here.
E
E
E
E
E
E
B
I
think
we,
we
also
have
a
small
reprieve
here
in
terms
of
the
overall
timeline,
the
open,
the
tree
project,
I
Sergei
posted
an
update
to
the
milestones
I'll
try
to
to
disseminate
this
more
clearly
to
everyone
so
that
they
have
it,
but
I
can
post
I'll
just
slide
this
into
our
agenda
doc.
Basically,
the
same
thing.
B
B
So
we
we
had
like
a
check-in
this
week
about
like
project
velocity
and
like
what
we
think
is
realistic,
and
there
is
a
desire
to
get
a
little
more
organized.
What
got
proposed
was
where
the
current
version
of
the
spec
is
called.
You
know
version
point
1
we'd
like
to
kick
the
alpha
off
with
a
couple
of
the
final
specs
that
are
in
flight,
namely
metrics,
and
this
kind
of
named
tracer.
B
You
know
component
thing
that
people
have
been
wanting
so
we'd
like
to
get
those
those
RFC's
merged
in
by
10
for
week
and
we're
gonna
version
that
spec
and
call
that
V
point
okay
and
then
try
to
ask
that
all
of
the
SIG's
that
are
like
at
this.
You
know
almost
there
kind
of
deadline
by
10:14
to
be
able
to
say
that
they're
up
to
like
the
point
of
the
spec,
which
I
believe
we
should
be
able
to
hit
without
a
problem,
because
we're.
B
We're
going
to
try
to
have
a
more
public
messaging
around
10:14,
saying,
like
hey,
open,
telemetry
is
in
alpha.
You
know
everyone
come
try
to
do
it
and
we
want
to
have
that
messaging
happen.
When
you
know
at
least
a
number
of
these
figs
are
hitting
that
deadline,
and
since
the
current
spec
is
called
V
point
1
we're
gonna
call
that
V
point
2,
so
we're
gonna
say
the
first
outflow.
B
We
release
is
like
the
point
2
version
of
the
project,
ok
and
then
there's
still
some
remaining
work
that
we
know
there's
a
set
of
specs
that
are
still
in
flight.
Some
unresolved
issues
around
global
initialization
context.
They're,
like
you,
know,
Hotel
protocol
semantics
conventions,
and
so
we
want
to
get
those
RFC's
completed
and
we'll
call
that
once
those
are
in
will
we'll
call
that
release
of
point
3
release.
Ok
and
of
course
this
is
all
negotiable.
This
is
just
kind
of
like
ok,
yeah.
E
E
B
B
B
So
there
is
actually
a
question
around
versioning
we're
discussing.
It
would
be
great
if
the
projects
were
able
to
match
the
version.
Number
of
this
spec
totally
sure
how
realistic
that
is,
but
I
did
want
to
check
in
if
we
could
be
moving
towards
a
versioning
scheme
where
we
were
saying
we
are
at
point
two
and
then
using
the
patch
number
is
like
our
build
or
patch
releases
to
the
project.
A
B
D
B
E
B
B
Exists
of
what
basically
exactly
I,
don't
impractical
ear
to
me.
What
was
problematic
about
this
approach,
but
we'd
like
to
try
something
where
there's
some
correlation
between
versions
were
using
in
the
SDKs
and
the
version
of
the
spec
that
it
actually
another
approach
might
be
to
say
that
you
know
that
if
we
have
in
our
readme
or
when
we're
saying
like
for
each
if
we're,
if
each
API
or
each
part
of
the
project
is
maybe
conforming
to
a
different
part
of
the
spec
I
could
see
some
way
of
communicating
that
yeah
but
yeah
anyways.
E
E
Don't
know
whether
you
can
see
it
or
not.
So
in
with
the
current
HTTP
plug-in
when
the
request
is
made,
we
get
this
band
for
getting
the
connection
DNS
the
actual
connection
sand,
so
there
are
like
four
different
span
created
for
the
plug-in
now
that
works
nice
for
the
demo,
but
in
reality,
if
somebody
were
to
use
the
plug-in
I,
don't
think
they
want
separation
of
all
the
spans.
There
will
be
too
many
spans
for
every
request,
so
I
believe
that
we
need
to
really
use
this
thing.
E
B
So
in
open
tracing
for
these
smaller
details,
we
had
the
open
tracing
log
interface,
which
I
think
in
open
census
was
like
timed
events
here,
like
that
and
so
yeah
to
degree
to
which
you
need.
You
want
to
know
about
these
operations,
but
it
seems
like
it
might
be.
Just
timed
events
or
something
like
that.
Yeah.
E
B
E
E
D
If
you
pull
up
the
issue,
I
asked
it
there
too,
but
basically
there
are
two
routes
we
can
or
kind
of
two
main
routes
we
can
go
with.
The
plug
in
one
is
what
the
open
tracing
plugins
did.
The
other
is
what
open
census
did
so
it
basically,
we
can
use
either
interceptors
or
stats
handler
stops
handler,
exposes
things
like
wire
lengths
which
are
useful,
but
it
doesn't
let
us
it
doesn't.
Let
us
capture
things
per
per
message
sent
in
a
stream.