►
From YouTube: 2021-11-01 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
Good
morning,
everyone
we'll
get
started
at
four
minutes
after
the
hour.
A
All
right,
let's
get
started
hello,
everybody,
starting
with
the
sig
check-in,
do
we
have
anything
for
the
non-metrics
part
of
the
spec,
because
we'll
cover
metrics
after
that
see
if
we
have
carlos,
yet
okay,
we'll
move
on
so
metrics
we've
had
some
big
releases.
I
was
wondering
riley.
If
we
have
riley.
If
you
want
to
go
over
it.
B
The
metrics
sdk
spike
has
reached
a
feature
freeze
and
ask
for
each
language
maintenances.
If
you're
working
on
the
matrix
implementation
in
your
language,
please
go
and
provide
feedback
by
creating
github
issues
before
we
went
to
stable
release
and
there
is
already
a
feature
matrix
created.
So
if
you
have
like
in
progress
work,
please
keep
the
metrics
up
to
date,
and
the
last
thing
is:
please
create
the
milestones
for
your
language,
where
you're
going
to
release
beta
rc.
A
And
riley
one
thing
that
might
be
worthwhile
talking
to
maintainers
about
so
we
have
we've
had
feature
freeze
for
the
main
metrics
specification.
We
feature
freeze
now
for
the
sdk
specification.
A
This
means,
like
you,
said
that
we
need
the
various
sdk
developers
to
you
know,
set
those
target
dates
and
start
implementing,
and
we
need
to
know
quickly
if
they're
going
to
run
into
run
into
problems
that
require
spec
changes
beyond
that.
What
else
do
we
have
sort
of
in
the
pipe
for
metrics
right
like?
Where
are
we
otlp
and
other
components.
B
So
the
protocol
has
reached
the
stable
a
few
months
ago,
so
it's
already
there.
I
I
I
think,
for
now
the
language
maintainer
should
focus
on
the
current
matrix
spike.
Okay,
just
heads
up
after
we
release
the
stable
version.
There
are
a
couple
of
things
we'll
add
in
the
next
stable
release,
so
things
like
the
base
two
exponential
histogram
and
a
more
like
highly
optimized
bond
api
loader.
Since
currently
we
track.
If
you
go
to
the
open,
telemetry
top
level
product
tracking,
you
can
see
there's
a
the
matrix
after
stable
release.
B
A
Perfect
thanks
riley
any
questions
from
maintainers.
I
know
this
is
like
the
first.
We
went
through
a
rather
large
process
last
year
for
for
tracing
and
then
earlier
this
year.
So
any
any
questions
for
maintainers,
because
this
is
a
big
milestone
for
the
spec
and
it's
a
big
chunk
of
things
that
we
need
to
now
implement
across
all
the
languages.
B
C
I
have
a
quick
question
regarding
two
features
that
were
removed,
particularly
batch
observers
and
the
bound
and
unbound
labels
for
languages
that
have
already
implemented
those
features.
They're
not
in
the
feature
freeze
spec.
Should
we
remove
them
or
should
we
keep
them
with
the
anticipation
that
they'll
be
added
back
into
the
spec
in
the
future?.
B
I
think
it's
likely
they
will
be
either
the
back,
especially
the
bundle
api
for
the
batch.
There
are
different
ways
of
achieving
the
same
goal.
My
suggestion
would
be
if
you
can
put
those
logic
in
a
separate
package
or
have
some
way
to
mark
that
as
experimental,
so
it
won't
affect
your
stable
release.
Then
you're
good.
I
I
think
removing
things
is
a
lot
of
effort,
so
one
thing
I
can
share
is
in
c
plus
plus
people
just
added
the
compile-time
micros.
B
A
Right
thanks,
riley
next
up
to
logs,
looks
like
we're
freezing
the
data
model
in
q1,
aiming
to
resolve
the
majority
of
open
items
in
q4
and
the
sig
is
now
switched
to
weekly
dan
or
anyone
else.
Any
updates
on
this.
B
A
A
E
Sure
so
the
two
we.
A
E
We
were
capturing
the
initially
we
were
capturing
all
span
attributes
which
of
course,
is
not
respect.
Actually,
neither
of
the
issues
that
came
up
were
they
were
both
because
we
didn't
follow
the
spec.
Initially
we
were
capturing
all
spam
attributes,
which
of
course
was
horrible,
and
then
we
found
last
week
through
a
user
reported
and
then
josh
also
through
his
testing.
E
He
ran
into
this
where
we
were
capturing
netpure,
host
and
netgear
ip
for
server
metrics,
which
of
course
is
very
high
cardinality,
and
that
was
because,
when
reading
the
metrics
http
metric
spec
table,
I
missed
the
column
that
differentiated
between
client
and
server
metrics
and
so
captured.
Netpure
host
for.
E
Yeah
yeah
we're
just
conforming
to
the
correctly
to
the
spec,
and
I
think
that
the
goal
eventually
is
to
my
understanding
from
talking
with
josh,
is
to
for
the
metrics
sdk
to
have
the
like
views
and
to
pat,
so
that
we
should
pass
all
the
span
attributes
to
it
and
it
will
do
the
filtering.
E
But
until
then
we
need
to
be
careful
and
do
that
before
passing
them.
Okay,.
A
A
A
Okay,
go
released
1.1.0
last
week,
active
work,
ongoing
for
metrics
logging
and
bug
fixes
c,
plus,
plus
logging
added
the
otlp
http
log
exporter
work
on
going
for
otlp,
grpc
log
exporter.
Logging
support
is
experimental,
of
course,
and
behind
the
feature
flag,
great
metrics,
initial
pr
created
for
the
metrics
api
reviews
in
progress.
Great
looks
like
no
updates
for
ruby
or
swift.
Do
we
have
any
collector
updates.
A
I
don't
know
if
I
see
anyone
from
the
collector's
sig,
so
I'm
going
to
assume
not
rust,
no
updates,
erlang
working
on
final
issues
for
tc
review,
great
and
rust.
Is
there
twice
all
right
moving
on
to
the
main
topics,
riley?
Is
anyone
interested
in
the
julia
programming,
language.
B
Yeah,
so
a
developer
reached
out
to
me
asking
some
metrics
spike
issue
and
I
realized
he's
working
on
a
project
basically
duplicating
the
language
sake
on
his
own,
and
I
talked
with
him.
It
seems
he's
very
passionate
about
julia.
He
doesn't
have
a
full
commitment
and
he's
looking
for
other
developers,
so
I
have
two
questions
number
one:
should
we
invite
him?
I?
I
can
work
with
him
to
start
the
julia
language
sid,
but
my
fear
is.
I
I'm
not
even
sure
how
many
developers
would
be
involved
like
whether
we
can
even
start
a
weekly
meeting.
B
B
Number
two,
I
think,
even
without
other
developers,
it
might
be
a
good
idea,
because
I
I
see
the
rights
in
the
language
based
on
my
discussion
with
him
and
it
seems
getting
popularity
in
the
big
data
processing
part.
I
I
think
it
might
be
better
if
he
could
contribute
to
the
open,
telemetry
project
that
helped
to
attract
other
developers
instead
of
having
him
working
on
some
separate
project
forever.
A
I
mean
I
think
that
would
generally
be
good,
obviously
it'll
be
challenging
if
it's,
if
it
ends
up
being
just
him
or
mostly
him,
who's
working
on
it,
but
I
think
that's
probably
better
than
it
sitting
off
on
its
own
and
that'll
also
be
a
way
for
him
to
come
into
the
community
and
learn
more
about
the
this.
A
Any
changes,
the
specification
things
he
needs
to
keep
track
of
and
finally,
presumably
there
are,
you
know,
there's
not
presumably
there's
there's
more
members
of
the
community
or
potential
members
of
the
community
than
there
are
people
than
the
18
of
us
on
this
call
right
now,
and
so
there's
there's
got
to
be
other
people
who
are
interested
in
julia
and
hotel,
even
if
they're
just
going
to
make
some
some
sort
of
fly-by
contributions
yeah.
This
will
be
a
way
to
capture
that.
B
E
So
I
would
probably
pause
on
starting
the
repo
structure
until
you
have
support
from
a
developer
community.
Okay,.
E
That
is
really
important
for
the
health
of
the
project.
I
agree.
The
adoption
is
really
important
that
we
have
language
support,
but
if
you
have
a
single
developer
working
on
something,
that's
not
good
that
there's
no.
E
There's
no
collaboration
across
different
companies,
which
is
one
of
the
the
main
points
that
this
project
was
founded
and
making
sure
that
there's
checks
and
balances,
I
think,
is
something
that
is
really
important
as
well
as
I
think
that
it
also
produces
better
code.
When
you
have
a
developer
community,
I
I
would
love
to
see
this
added,
but
I
do
not
want
to
see
it
added
if
it's
a
single
developer,
because
one
it
would
disenfranchise
them
or
them,
and
it's
not
a
really
good
community
interaction.
E
So
I
I
would
definitely
would
like
this
to
be
added,
but
I
would
like
to
be
a
have
some
sort
of
commitment
level.
You
know,
even
if
that
commitment
level
is
like
three
different
companies
say
that
they
can
commit.
You
know
one
developer,
you
know
five
days
or
one
day
a
month,
even
like
it's
something
rather
than.
E
A
single
developer
push
this
because
otherwise
it's
equivalent
to
having
them
host
it
and
then
us
putting
it
in
some
sort
of
like
registry,
or
something
like
that.
I
I
think.
B
I
work
on
the
c
plus
plus
at
that
time
it
was
just
a
repo,
even
without
a
complete
readme,
and
I
started
that
then
I
put
two
developers
from
microsoft
and
at
that
time
it
was
basically
developers
from
microsoft
writing
code,
whatever
they
want
and
later
we
have
other
folks,
because
the
product
seems
to
have
progress,
then
the
other
folks
start
to
join
and
we
refactor
a
lot
of
things
and
we
rewrote
a
lot
of
things
just
to
make
sure
it's
well
aligned
and
now
we
see
the
stable.
B
So
I
have
the
gut
feeling
that
is
a
chicken
egg
problem,
and
I
agree
with
your
tyler
like
like
there
there's
a
important
thing
like
we
don't
want
something
to
just
sit
there
and
nobody
contribute
to
it.
B
But
I
I
guess
the
stepping
back
question
is
when
we
face
this
chicken
egg
problem.
Do
we
want
to
be
a
little
bit
aggressive
or
we
want
to
be
conservative?
I'm
happy
either
way,
I'm
trying
to
be
helpful
here
and
I
think
for
julia
case.
It's
more
like
the
language
is
not
mature
enough
and
it's
still
evolving.
So
I
think
in
this
way
it's
more
like
the
education
part.
If
we
can
make
this
available
when
we
attract
other
developers,
it
might
set
a
good
starting
point.
E
B
D
Oh,
I
was
gonna
say
I
I
have
like,
I
think,
two
concrete
suggestions
and,
and
one
word
of
warning,
which
is
like
this
in
open
tracing.
There
were
a
couple
cases
where
we
were
pretty
quick
to
move
a
project
in
when
it
was
like
this
and
had
a
single
developer
like
lua.
This
is
an
example
of
this
and
it
didn't
necessarily
end
well
for
a
lot
of
the
reasons.
D
Tyler
was
saying,
I
think
one
one
thing
that
I
think
is
like
a
good
middle
ground
step
is
we
do
we
ask
the
tc
to
do
audits
before
something
you
know
goes
1.0.
D
I
I
think
before
moving
this
in
a
great
way,
to
maybe
be
encouraging
would
be
to
to
just
audit
the
api
in
particular
and
just
verify
how
how
on
track
it
is
relative
to
the
spec
and
and
give
them
notes
and
and
see
what
they
do
with
those
notes,
and
if
that
process
seems
like
it's
going
well,
that
would
be.
That
would
be
a
good
sign
that
this
might
be
a
good
fit
to
become
an
official
repo.
D
So
so
I
think
that's
like
a
good.
There
should
perhaps
be
a
feeling
out
phase.
I
guess
that's
what
I'm
saying
yeah,
where
we
verify
we
can
collaborate,
one
of
the
cases
I'm
thinking
from
the
past
was
we
said
yes
to
this
project
and
then
tried
to
send
that
maintainer
notes
and
they
were
like.
Well,
I
don't
want
to
do
those
things
I
said
no,
not
like,
and
so
so
that
was
awkward.
D
You
know
so
yeah
one
thing:
I've
wondered
when
people
have
come
to
me
asking
about
pearl
and
some
of
these
other
languages
is
how
feasible
is
it
in
that
language
community
to
bind
to
c
plus,
because
that
to
me
seems
like
if
that's
not
awkward
or
or
crazy
for
that
community,
that
might
be
a
good
good
middle
zone?
Maybe
that's
not
what
this
person
wants,
but
yeah,
if
it's
possible
to
focus
the
efforts
of
that
group
more
on
like
the
api
and
the
instrumentation
and
stuff
and
just
bind
to
an
existing
implementation.
B
B
The
second
part
is
maybe
we
can
do
something
like
a
staged
approach,
so
for
people
who
want
to
start
some
project,
they
can
start
as
a
very
early
experimental
version
and
as
it
like,
we
can.
We
can
have
a
some
like
status
for
each
language
sake,
whether
the
community
is
still
exploring
similar
like
cincia.
If
we
have
the
like
multiple
stages,
whether
it's
incubation
or
sandbox,.
D
B
D
That
that
might
I
it
sounds
like
we
might
gain
some
speed
now,
but
it
could
actually
block
progress
in
the
future.
Just
I've
kind
of
seen
that
happen.
A
D
I
think
it's
it's
partially
just
just
just
getting
to
know
the
the
maintainer,
especially
if
it's
just
one.
D
A
The
only
maintainer
for
php,
so
it
it
can
it
it
can
work,
though
it's
it's
a
huge
commitment,
as
I
think
bob
could
probably
attest
to
yeah.
D
D
E
E
I
like
the
idea
that
ted
proposed,
where
we
have
some
sort
of
review
process
for
like
what
needs
to
be
addressed
and
things
that
need
to
be
like
brought
in
and
if
you
see
activity-
and
you
see
a
community,
that's
starting
to
sprout,
to
try
to
support
that.
I
think
that's
a
really
good,
you
know
clearance
to
say
like
yeah,
okay,
like
now,
let's
bring
this
in
house.
I
like.
E
I
really
I
really
like
the
idea
of
leaving
it
external
until
you
can
get
some
sort
of
commitment.
I
think
that
that's
a
positive
step
and
like
there
are
ways
that
we
can
advertise
different
languages
and
different
language
supports
that
are
technically
unofficial.
We
do
that
with
the
registry.
I
don't
know
if
we
could
do
that
in
a
better
way,
but
I
think
that
having
the
experimental
stuff
live,
external
is
a
really
great
way
to
to
provide
creativity
without
having
to
you
know,
commit
ourselves
to
supporting
julia
for
the
next.
F
A
B
A
A
F
D
B
D
B
C
B
A
Yeah
self-fulfilling
prophecy
he's
in
riley
he's
in
asia,
yeah,
yeah,
okay
and
then,
as
for
having
them
chat
with
the
tc,
do
we
want
to
have
a
tc
champion,
or
at
least
one
tc
member?
They
should
reach
out
to
first
carlos.
I
see
her
on
the
call.
Can
that
be
you.
A
D
C
D
There's
tc
group,
I
think,
it'd
be
good
to
post
that
there
I
I'd
like
to
to
encourage
the
tc
to
to
organize
themselves
and
not
just
get
okay.
C
A
Okay,
next
topic,
john:
I'm
changing
teams
at
splunk
and
will
be
reducing
my
involvement
in
the
project.
John
says
more
java
contributors
are
needed,
especially
to
for
careful
pr
reviews.
F
Yeah,
so
I
just
wanted
to
you
know,
make
sure
other
maintainers
are
aware:
I'm
going
to
be
I'm
changing
teams
still
working
observability,
obviously,
but
I'm
going
to
be
stepping
my
involvement
down
a
little
bit
probably
will
probably
not
be
going
to
spec
meetings
just
from
a
time
per
time,
commitment
perspective,
and
we
really
really
need
more
people
to
do
very
careful
pr
reviews
in
java.
So
if
there's
people
with
your
companies-
or
you
know
who
are
interested
in
doing
that,
please
encourage
them
to
do
so.
F
Was
it
something
we
said
it's
two
years
of
being
a
maintainer
is
exhausting.
Yes,
that's
what
it
is.
I
mean
this
is
to
call
out
what
we
were
talking
about
previously,
I'm
I'm
just
burned
out.
I
need
a
break
fair
enough,
I'm
going
to
continue
as
a
maintainer,
but
I'm
not
going
to
be
quite
as
involved
and
mostly
just
kind
of
provide
support
for
honor.
At
this
point.
A
A
Okay,
any
questions
comments
concerns
okay,
tristan
I'm
working
on
some
proposals
for
spam
processors.
I
plan
to
just
look
into
each
language's
implementation
emphasis
on
what
the
spec
calls
decorators
to
see
what
they
support.
If
you
know
of
any
plans
for
your
language
that
won't
be
found
in
docs
or
code.
Yet
please
cling
me
on
slack.
C
I
have
a
quick
question
just
to
see
how
other
maintainers
have
handled
a
situation.
The
specification
for
propagators
specifically
for
the
environment
variables
says
that
the
jaeger
remote
sampling
propagator,
like
it,
defines
that
as
as
a
potential
value,
but
nowhere
in
the
sdk
specification
is
that
propagator
specified
or
required,
or
anything
like
that.