►
From YouTube: 2022-04-07 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).
C
B
A
Yeah
right
portland's
supposed
to
get
to
76
today,
which
is
warm.
I
think
we're
going
right
back
to
that.
I
think
highs
of
like
50s
or
something
like
that
so
good
day
to
be
outside
instead
of
at
a
computer.
A
Yeah,
I'm
not
either
sounds
like
anthony
might
be
going
we'll
have
some
representation.
B
C
A
B
I
mean
there's
seattle.
Seattle
was
great,
there's
been
washing
san
diego
here
in
austin's
hosted
one,
but
detroit.
A
A
A
C
A
I
didn't
have
too
much
to
talk
about.
I
just
wanted
to
talk
about
like
releases
and
stuff
going
on,
but
you
could
probably
wait
well
we're
about
to
after.
A
I
know
right,
it's
not
yeah.
I
was
just
complaining
about
not
being
outside
so
right,
yeah,
but
yeah
we
can.
We
can
just
jump
in
here
see
david's
on
the
call
as
well
hey
if.
A
To
the
attendees
list,
I
think
you're,
the
only
other
one
on
the
call
they
can
jump
in
here
and
start
sharing.
A
Okay.
I
think
this
is
the
right
one:
okay,
cool
yeah,
so
cool
yeah-
it
was
the
only
other
thing
I
was
thinking
about
is
metric
update
the
release
status,
I'm
guessing
everyone
on
the
call
saw
that
we
got
out
1.6.2
yesterday,
thanks
to
anthony
watching
this
there
is,
there
were
some
compatibility
issues
with
the
collector
and
the
1.6.1
upgrade
used.
I
think
it's
v012,
I'm
not
the
most
current
on
the
compatibility
issues
here,
but
it
used
v0.
A
I
think
12
of
the
protobuf,
and
there
was
compatibility
issues
with
the
recent
collector.
It
sounds
like
anthony
he's
been
digging
into
it
here,
and
this
is
something
to
I
think
read
the
today
is
there's
more,
but
there
was
an
issue
with
dropped
spans
with
some
combinations
of
the
collector
and
the
otlp
that
we
were
shipping
with
v01rp161.
A
I
still
think
there
are
probably
some
compatibility
issues
if
you're
using
a
really
old
collector
it
sounded
like.
So
I
don't
know
if
this
is
completely
resolved
from
the
the
user's
perspective.
I
think
it
might
be
resolved
from
our
perspective.
It's
to
like
all
do.
The
only
other
option
was
there's
a
a
pr
that
was.
A
I
think
this
actually
is
the
pr
to
emit
both
the
instrumentation
library
spans
and
the
span
scopes,
essentially
a
duplication
of
the
data
which
I
think
is
a
misstep,
and
I
think
it
was
recognized
in
here
because
the
the
wire
compatibility
essentially
you're,
adding
on
an
additional
payload
fields.
So
that's
probably
not
what
we
want
to
do,
but
we
do
want
to
make
sure
that,
like
at
least
understand
the
compatibility
issues
and
maybe
communicate
those
in
the
documentation
or
something
like
that,
I
think
it's
probably
a
good
idea,
but
oh
just.
B
I
did
a
little
bit
of
exploring
with
this.
I
didn't
get
to
do
too
much,
but
if
you're
using
161
or
before
and
otlp
015.
B
You
will
be
sending
out
on
the
new
message
type
that
was
created
just
because
of
how
names
work.
So
if
you
use
a
collector
before
48,
it
will
just
not
understand
it
whatsoever.
B
I
thought
version
48,
oh
48
o48
should
take
the
new
message,
types,
the
new
instrumentation
library,
which
is
now
spot
1000
in
the
in
the
protocol
and
convert
it,
but
it
may
or
may
not
be
doing
that.
So
we
just
need
to
be
aware
of
that.
B
But
if
you're,
using
161
with
otlp
prior
to
o15
or
you're,
using
162
with
proto
o15,
both
47
and
48,
should
should
be
accepting
the
messages,
because
the
the
wire
format
hasn't
changed
under
the
hood.
A
Okay,
so
I
mean,
I
think
that
our
our
community
recommendation
is
to
upgrade
just
go
to
the
latest
for
all
of
them
yeah
and
I
think,
we'll
be
able
to
hit
our
targets
there,
but
yeah.
There
is
some
weird
compatibility
issues
as
we
kind
of
anticipated
with
this
change
but
yeah.
I
think
I
think,
for
all
of
the
people
who
are
experiencing
this
we've
identified
in
and
I
think
the
path
forward
is
clear
for
them
in
the
upgrade
cool.
A
So
that's
one
bug
the
next
one.
It
was
I
think,
identified
this
morning
that
the.
A
Tracer
provider
or
the
text
maps
propagated.
That's
what
it
is.
If
you
try
to
set
the
propagator
twice,
what
it
will
do
is
that
it
will
actually
do
a
comparison
against
two
potentially
non-comparable
types.
A
So
like
the
new
or
a
composite
text,
micropropagator
is
not
comparable
the
first
time
it
doesn't
really
matter
which
also
by
the
way,
I
learned
that
today,
if
you
try
to
compare
a
comparable
and
a
non-comparable
it'll,
it
won't
panic,
but-
and
it
will
tell
you
it's
just
not
equal
or
equal
because
it
can
tell
it
obviously
isn't
once
comparable.
The
other
isn't.
But
if
you
try
to
compare
two
non-comparables,
that's
where
a
panic
is
thrown,
and
so
that's
why
setting
this
twice
is
the
problem.
A
We
took
the
validation
outside
of
the
do
once
protocol,
and
so
because
of
that,
we
it
results
in
this
essentially
working
to
try
to
do
a
comparison
twice.
A
So
that
said,
there's
a
patch
here
and
looks
like
it
has
the
approvals
to
fix
all
that.
Essentially,
you
we
do
a
type
assertion
for
any
sort
of
comparison
to
ensure
that
the
type
of
comparing
is
comparable,
which
we
know
because
we
control
this
type
for
both
the
or
for
all,
the
trace
or
provider
media
provider
and
the
text
propagator.
A
So
I
was
I'm
hoping
to
push
this
out
today
again
because
having
a
panic
in
something
it's
it's
kind
of
an
edge
case,
but
it
also
is
not
a
good
look
so
yeah.
I
I
think
that
I
want
to
try
to
get
this
out
soon.
I
will
probably
merge
this
after
the
call
given
it's
got
the
two
approvals.
If
anybody
on
the
call
wants
to
still
review
this,
please
just
speak
up
and
I'll
wait
to
merge
this
before
you
take
it.
Take
a
look.
A
A
A
That
being
said,
the
only
thing
then
left
is
we
had
another
feature.
Release
scheduled,
and
this
is
hasn't
really
changed.
In
the
past
week
I
tried
taking
a
look
at
the
in-memory
metrics
exporter
and
got
distracted
by
20
other
things,
so
I
didn't
make
it
too
far
on
that
one,
but
otherwise
I
think
that
we
need
to
well.
A
I
wanted
to
get
some
feedback
on
this
issue
here
or
this
pr
here
I
created
a
way
to
generate
the
rest
of
the
semantic
conventions,
targets
which
is
nice,
because
otherwise
we
were
just
copying
code
and
propagating
errors,
or
you
know
things
of
that
are
prone
to
errors
from
humans.
A
So
this
essentially
takes
all
that
generation
and
moves
it
to
a
make
target,
and
you
run
the
big
target
and
it's
reproducible.
It
is
kind
of
daunting
how
big
it
is.
I
I
want
to
make
sure
that
I
that's
not
an
inhibitor
for
people
who
are
viewing
this
one
of
the
things
that
you
know.
There
are
a
lot
of
points
here.
I
could
split
this
up.
A
It
is
going
to
take
a
little
bit
longer,
but
I
just
wanted
to
like
see
if
there's
any
feedback
as
to
what
I
could
do
to
help
progress
this.
So
we
can
now
generate
the
four
versions
of
some
comp
that
we
are
behind
on.
A
Okay,
yeah-
and
I,
like
I
said,
there's
a
lot,
but
you
know
most
of
the
red
is
just
the
fact
that
in
this
pr,
what
I'm
doing
like
there's
a
tool
here,
which
is,
should
be
really
straightforward
to
review.
But
then
a
lot
of
the
pr
is
just
template,
stuff,
which
we've
already
reviewed,
because
it's
just
the
same
code,
with
the
big
caveat
being
that
I
needed
to
abstract
the
http
libraries.
A
So
this
one
is
a
little
bit
a
little
bit
of
a
change
in
the
fact
that,
like
it
essentially
wraps
semantic
conventions
into
this
internal
type
and
then
from
that
internal
type,
it
can,
you
know,
expose
all
the
functionality
that
we
had
before.
I
think
that's
where
the
majority
of
the
review
is
going
to
come
in,
so
just
a
heads
up
on
on
what
you're
expecting
when
you
see
4
000
lines
of
deletions.
It's
mostly
just
taking
all
of
this
code
and
replacing
it
and
then
the
tests
are
all
deleted.
A
A
Yeah,
I
think
with
that,
is
all
the
release
news.
The
the
go
sig
newsletter
is
reviewed
at
this
point.
Yeah
so
I'll
hand
it
over
to
aaron
get
a
little
update
on
the
metrics
progress
and
yeah.
I'm
interested
in
this
one.
B
B
B
This
is
kind
of
in
contrast
to
how
prometheus
does
its
collections.
It's
you
know,
prometheus
knows
when
to
trigger
a
collection
based
on
you
know.
The
http
request
versus
a
otlp
exporter
has
no
idea
when
to
collect,
so
we
need
some
something
to
kind
of
bridge
that
that
gap
and
the
periodic
exporter
or
the
periodic
exporter,
I
guess,
is
names.
Names
are
hard,
it's
either
an
exporter
or
a
reader
that
that
is
what
will
fill
that
gap.
B
The
next
thing
is
test
coverage.
A
number
of
tests
were
removed
due
to
things
changing
in
those
areas,
and
that
needs
to
be
brought
back.
I
think
we
only
have
about
40
to
50
coverage,
we're
where
we
were
up
in
the
like
70
80.
Prior
to
this,
so
we're
gonna
bring
a
lot
of
that
back
and
then
the
the
final
step
is.
We
want
to
kind
of
reorganize
this
and
rebase
it
so
that
it
a
can
be
merged.
B
It's
not
in
a
murderable
state,
because
it's
not
been
rebased
onto
the
latest
main,
but
also
to
give
a
better
story
for
how
things
are
progressed.
Sort
of
in
a
similar
vein
to
how
the
api
was
was
split
into
a
number
of
different
steps
in
in
apr
process.
I
we
want
the
commits
to
kind
of
tell
that
same
story
where
the
first
step
is
removing
everything.
The
second
step
is
laying
the
base
layer
of
stuff
fixing
the
the
tests
that
are
external
to
the
sdk.
B
Like
you
know
the
test
in
exporters
and
and
in
examples,
and
that
way
we
can
actually
have
like
a
lot
more
intelligible
reviews,
because
at
some
point
it's
still
gonna
be.
You
know:
minus
five
thousand
plus
five
thousand,
but
if
you
have
it
in
steps,
it's
it's
easier
to
kind
of
digest.
The
flow
of
things.
A
Yeah
that
sounds
good.
I
appreciated
the
commit
structure,
the
last
time
helped
and
just
looking
at
it
in
that
platform.
So
yeah,
I
think
that's
that's
useful.
I
also
think
if
you
can
leverage
what
we
did
with
the
globals
package,
where
essentially
like
we
counted
out
or
put
to
do's,
then
we
have
some
issue
tracking.
It
was
a
really
useful
process,
because
we
can
also
not
only
progress
the
main
issue
faster
but
distribute
the
workload,
especially
for
that
kind
of
stuff.
B
Yeah
seasoned
members,
there
there's
an
issue
there
in
that
we
are
still
kind
of
on
the
edge
of
we
need
to
make
sure
that
our
core
set
of
logic
is
working
versus
the
the
like
the
external
exporters.
Those
might
actually
stay,
as
commented
out
in
the
in
the
the
review
version,
but.
A
B
Yeah
and
we're
still
in
the
process,
one
of
the
things
that
I
wanted
to
bring
up,
or
at
least
mention,
is
we're
adding
tests
internal
into
the
sdk
portion
of
it
just
to
make
sure
that
our
core
logic
is
right
and
we're
still
finding
like
very,
very
hard
edge
cases
to
deal
with
where,
like
duplicate
instruments
or
an
instrument,
is
in
multiple
callbacks
or
something
like
that.
How
do
you?
A
Yeah,
I'm
glad
she's
involved
in
that,
because
that
has
been
a
topic
of
mixing
for.
I
guess
I
can
say
years
now.
Yes,.
B
So
yeah
it
feels
like
decades,
but
not
quite
there.
Yeah
in
software.
A
Terms,
I
think
it
is
decades
so
yeah,
oh
cool
yeah.
That
all
sounds
really
good.
Thanks
for
the
status
update.
I
know
it's
a
lot
of
work,
so
I
appreciate
it
cool,
I
think
that's
it
for
all
of
the
agenda
items
unless
other
people
had
something
they
wanted
to
talk
about
or
bring
up.
A
Okay,
yeah
small
group,
small
agenda,
I'm
hoping
everyone's
out
enjoying
the
weather
wherever
the
weather
is.
D
A
Cool
yeah-
again,
that's
understandable
it
sometimes
I
personally
it
took
a
while
for
me
to
find
something
to
work
on
when
I
first
started
so
yeah
totally
good.
A
D
I'm
working
on
with
the
helm,
charts-
and
I
don't
after
I'm
done
with
that
I'll,
be
I'll,
be
ready
to
pick
something
up.
So
just
gotta
finish
that
first
yeah.
A
Definitely
appreciate
it
thanks
so
much
okay
cool,
then
I
would
like
to
end
it
here.
I
think
I
don't
think
we
have
anything
else
so
yeah.
I
appreciate
everyone
taking
the
time
and
coming
and
chatting
everyone
watching
the
video.
I
guess
also
thank
you
we'll
same
place
same
time
next
week
see
you
then,
and
otherwise
online
slack
or
issues
till
then
bye.