►
From YouTube: 2021-02-25 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
B
C
B
C
C
So
the
first
agenda
item
that
I
had
for
today
was
as
part
of
open
telemetry
one
point
reaching
the
spec
reaching
1.0
the
powers
that
be
gave
us
the
option
to
switch
to
using
the
cncf
slack
rather
than
using
gator.
Does
anybody
have
a
strong
opinion
on
that?
My
strong
opinion
is
that
we
should
switch
to
slack
because
I
use
it
for
a
bunch
of
other
garbage
and
it
would
just
be
easier,
but
I'm
very
happy
to
stay
with
gitter.
If
anybody
has
a
strong
opinion
about
that.
C
Me
too,
I
think
I
figured
that
was
it's.
That
means
it's
just
one
command
click
away,
rather
than
switching
to
a
different
app,
but
I
will
so
what
I'll
do
is?
I
will
post
I'm
gonna
make
a
note.
I'm
gonna
I'll
post
instructions
on
how
to
get
signed
up
for
the
cncf
slack
in
the
channel
and
then
I'll
set
like
redirect
all
of
our
attention.
C
C
C
So
the
next
thing
that
was
on
my
agenda
was
ticket
prioritization.
I
realized
that
we
haven't
really
had.
We
haven't
really
been
good
about
prioritizing
the
tickets
that
we
have
for
this
project.
So
does
anybody
else
have
perspective
on
some
of
the
things
that
we
should
work
on
next
to
try
and
move
more
towards
our
general
availability?
I
know
benjamin
did
a
lot
of
leg
work
in
that
document
for
trying
to
see
where
we
fell
short,
I
feel
like
we
should
fix
some
of
those
things
as
our
higher
priority.
C
I
know.
Darlabor
also
did
that
open
telemetry
user
research
document
and
that
also
has
a
bunch
of
other
prior
other
pieces
of
prioritization
with
respect
to
tracing,
and
then
I
know
we
have
like
a
bunch
of
languishing
tickets
for
metrics
and
stuff,
but
I
think
that
at
least
from
my
perspective,
what
we
should
try
and
do
is
we
probably
try
and
focus
on
getting
tracing
completely
up
to
spec
first
and
then
start
slapping
through
metrics
a
little
bit
more
talk
to
you.
I
know
you're
working
on
the
metrics
part
in
that
part.
C
That
is
very,
very
helpful
because
otherwise
we
would
be
so
far
behind
in
that
realm,
but
I
think
that
what
that
we
can
we
can
keep
like
having
there.
There
are
a
couple
other
people
working
on
metrics
to
psych.
Maybe
we
could
keep
those
people
working
in
that
channel
and
then
the
rest
of
us
can
like
continue
to
finish
up
the
metrics
for
does
that
work
for
everybody.
C
C
Who
knows
all
right,
so
I
wanted
to
mention
that
we
have
spec.
Freeze
was
that
on
the
15th,
which
is
good.
They
also,
there
are
a
bunch
of
other
sig
subjects.
Groups
like
this
one
that
are
releasing
their
generally
available
1.0
soon,
they're,
all
the
ones
that
are
they're
going
to
be
in
like
no
particular
order,
but
they're
going
to
be
you
know,
go
is
really
saying
java
is
releasing.
C
B
Yeah
so
hi
guys,
I
mean
I
never
spoke
with
you
all,
but
I
have
been
chatting
with
bob
from
quite
a
while.
I'm
I'm
currently
working
with
atd
and
I'm
working
the
project
where
we
need
the
tracing
support
in
that
project.
We
were
earlier
using
open
tracing,
I
mean
in
the
open
tracing
php
library.
B
The
official
one
is
not
supporting
udp,
but
there
are
bunch
of
there
are
a
bunch
of
forks
been
taken
from
that
which
has
leveraged
on
the
library
and
they
built
the
udp
support
for
php
and
they're
like
four
five
or
six
out
of
there,
but
I
really
don't
want
to
use
one
of
those
because
I
mean
we
never
know
because
those
are
like
four
brands.
Somebody
will
work,
not
work,
it
will
get
upgrades
or
not.
B
So
that's
the
reason
I
I
thought
of
you
know
putting
my
effort
to
you
know
built
it
up
for
open
telemetry
php.
B
Even
I
spoke
with
bob
before
you
know,
working
on
this
thing
so
found
that
the
way
we
have
built
this
first
initially,
we
as
we
all
know
like
we-
are
supporting
zipkin
zipkin
http
right
and
which,
which
is
basically
going
to
send
out
traces
to
the
collector
right
and
the
collectors
collector,
is
accepting
the
traces
in
32k
hex
for
trace
id
and
16k
hex
for
the
span
id.
B
So
when
I
was
working
on
this
udp
thing,
I
found
a
little
bit
difficult
and
figuring
out
like
what
exact
agent
the
aegis
is
accepting,
which
format
agent
is
accepting
and
found.
Finally,
it
is
not
accepting
the
hex.
It
is
accepting
a
decimal
number,
which
is
the
integer,
sorry,
not
a
decimal,
but
an
integer
64-bit
number
and
which
then
again
internally
converts
it
to
the
hex.
What
collector
requires
and
then
send
that
trace
to
collector,
so
this
this?
Actually,
this
part
a
particular
thing.
B
I
think
we
are
the
one
who
who
has
actually
started
using
it
in
production,
and
we
found
the
found
some
of
the
grey
area
where
it's
our
all
the
traces
when
each
request
the
traces
are
getting
generated.
It
is
collect,
it
is
hitting
the
collector
server,
which
is
kind
of
odd,
so
normally
you
don't
and
there
and
collector
server.
It
is
again
going.
B
B
So
the
way
it
will
be
like
the
places
would
go
to
the
agent
and
then
agent
will
send
it
to
the
collector
in
batch
so
that
there
won't
be
direct
impact
on
the
collector
server.
So
that's
that's
the
reason
I
started.
Building
it
and
then
I
find
out
like
this
is
how
we
are
now
collector
is
accepting
hex
and
agent
is
accepting
the
64-bit
int.
B
So
we
have
to
find
a
way
so
that
I
mean
currently
the
pr
I
have
with
six
commits.
This
is
the
one
commit
last
commit.
B
I
have
done
just
now
just
before
the
call
which
is
kind
of
for
working
for
edgar
udp,
but
it's
breaking
the
httpg,
because
I
have
changed
the
data
type
to
integer
right,
so
we
I
actually,
I
have
to
work
on
this
to
find
out
like
a
midway,
so
it
will
support
both
zipkin
and
and
the
edgar
udp
for
the
when
we
have
when
you
are
passing
it
to
the
agent.
B
So
we
have
to
find
a
way
like
where
we
should
how
we
should
approach
to
this
problem
did
like
we
could
use
the
conversion
and-
and
I
I
found
another
way
actually-
I
try.
I
have
to
try
it
out
because
currently,
as
we
have
this
64-bit
int
already
there
in
the
pr,
I
was
trying
to
convert
it
to
the
hex
before
sending
it
to
the
collector,
when
the
call
is
happening
to
the
http,
so
that
kind
of
didn't
work.
B
Maybe
maybe
I
have
to
see
like
quite
why
it
was
not
working
but
yeah.
We
have
to
figure
it
out.
Yeah.
C
I
think
you
have
to
be
careful
because
the
tracing
spec
sp,
the
tracing
spec,
specifies
the
hex
character.
I
know
we've
looked
at
that
initially
with
the
integer,
but
you
can't
use
an
integer
because
of
that
hex
value
has
it's
ref
in
it
too.
So
that's
that's
like
that.
We
will
probably
have
to
find
some
other
approach
like
like
potentially
converting
it
at
some
other
point
in
this
stage,
but.
B
B
Yeah
yeah
yeah,
so
so
the
idea
behind
this
when
I,
when
I
was
working
on
this,
because
this
was
one
of
the
priority
because
our
servers
were
like
I
mean
it
was
in
age
and
we
could
not
push
other
projects
in
because
of
this,
because
if
other
process
other
projects
of
ours
will
started
using
this,
the
connect
collector
will
break
so
so
so
so
that's
the
reason
I
just
try
to
do
it
quickly
to
fix
these
things.
Things
up
then
later
to
optimize.
B
This
definitely
looking
forward
your
help
on.
You
know
to
optimize
this
and
then
see
how
what
should
be
used,
if
you
can
use
integer
and
then
convert
it
to
hex
before
sending
to
the
collector.
Is
this
a
good
way
or
you
can
just
stick
to
a
string
and
then
find
another
way
which
can
be
used
to
push
that
in
64
bit
into
the
agent?
B
There
are
two
ways
so
we
can
find
it
out,
maybe
in
a
call
or
maybe
two
or
maybe
in
chat
while
discussing
if
you
once
you
have
the
idea
after
looking
into
my
pr
what
to
do.
C
I
think
I
think
discussing
this
in
chat
would
be
good,
because
I'm
not
the
authoritative
source
on
this
problem
and-
and
there
are
a
lot
of
other
people
that
would
probably
like
to
weigh
into
it
as
well.
I
think
that
we,
I
think
that
we
just
have
to
be
careful
to
try
and
follow
the
spec.
We
also
have
to
be
careful
to
not
use
this
library
in
production.
Yet
the
first
thing
that
says.
C
B
I
mean
when
I
saw
the
open
tracing
php.
I
wonder
like
why
they
have
not
built
the
udp
thing,
because
that's
the
way
actually
to
go
to
the
production,
because
we
cannot
directly
hit
the
collector
and
it
will
blow
up
sometime
when
a
lot
of
requests
will
be
going
to
the
server
and
then
from
server
to
a
collector
server
right.
So
because
I
saw
the
languages
are
built
and
they
have
the
udp
thing
there
right
in
java
or
python
or
go
so
yep.
B
A
C
I
think
that
we
yeah
this
is
something
that
I'd
like
to
talk
with
more
than
just
me
about,
and
it
will
be
interesting
to
hear
your
experiences
from
bringing
us
like
putting
this
library
in
production
where
you're
seeing
zap
gaps.
It's
almost
like
we're
gonna
trace
the
tracing
utility,
but
we
yeah
we
like
get
our
feedback.
There
will
also
be
very
valuable,
especially
if
you
can
tell
us
where
it's
falling
short
from
a
performance
standpoint
we
may
have,
you
know,
may
have
a
memory
leak
or
a
cpu
spike
utilization
or
whatever
here
or
there.
C
B
Cool
cool,
so
I
mean
one
suggestion
like:
if
you
can,
if
you
have
some
time,
just
go
to
the
pr
and
let
me
know
like
where
else
we
can
improve,
maybe
do
we
need
to
stick
with
string
and
then
we
have
to
find
a
way
to
find
a
way
to
convert
it
to
hint
and
then
pass
it
to
the
agent
or
we
can
just
use
the
int.
And
then
we
have
to
find
a
way
to
convert
it
to
stream
extreme
before
sending
it
to
the
collector.
C
I
will
do
my
best
to
review
your
pr
chill.
Thank
you.
Last
but
not
least,
on
the
list.
We
have
our
internship
opportunity.
We
have
to
select
an
internship
candidate
before
the
deadline
is
the
26th
I'm
planning
on
as
of
right.
Now
I'm
planning
on
reviewing
all
of
the
candidates
either
tonight
or
tomorrow,
they're
all
on
the
cncf
page,
I'm
happy
to
share
that
link.
If
anybody
cares
about
it.
A
C
Don't
I
can
just
be
authoritative
and
pick
the
right
person.
I
have
a
person
in
mind
already
because
it
already
started
contributing
to
the
project,
but
I
will
I
just
didn't
want
to
make
that
authoritative
decision
without
like
getting
everybody
else's
input
as
well.
C
I
think
it's
from
march
to
june,
or
something
like
that
they
have.
They
have
like
three
segment
like
three
two
or
three
week
segments.
I
can
tell
you
one
second.
C
Yeah
it's
february
it
is
three
months
12
weeks,
full-time
schedule,
starting
on
march
1st.
D
C
Those
were
the
things
that
I
had
on
my
agenda
for
today
amber
from
family
journey.
Do
you
have
anything
that
you'd
like
to
go
over.
C
C
D
C
Yeah
I
think
there
I
know
I
can't
pronounce
this.
I
can't
I'm
not
very
good
at
pronouncing
this
name.
Ownium
said
he
was
gonna
work
on
documentation,
so
I'm
expecting
a
pr
for
him
from
him
shortly
since
he's
not
on
the
call.
That's
the
person
that
I'm
planning
on
offering
the
full-time
internship
too.
Oh
sweet
he's
done
some
work
already
and
I
think
he's
interested
in
the
project
and
that's
important.
D
Yeah
it's
funny
because
when
you
were
mentioning
interns-
and
I
was
thinking
documentation-
I
mean
not
that
that's
all
that
interns
are
good
for,
but
I
think
they
often
come
in
and
they
they're
really
in
a
learning
mode
and
like
yes,
we
need
somebody.
A
C
E
Cool
document
feedback.
C
It's
I
think,
it's
somewhere
deep
in
the
getter,
but
I
know
I
have
it
handy
in
my
open
club
in
my
google
inbox
or
my
google
drive.
I
think
this.
No
that's.
C
C
C
So
I
think
jody
that,
like
what
is
that
page,
three
and
four
in
that
document
or.
A
C
Three
through
seven
and
that
document
don't
have
associated
tickets
yeah
and
those
are
probably
the
ones
that
we
need
to
get
get
slapping
through.
First.
Okay,
that's
a
pretty
that
document's
a
pretty
good
gold
mine
for
all
the
things
that
we
need.
A
lot
of
the
things
that
we
need
to
complete
still.
C
Fami,
how
about
you,
you're,
you're,
still
slapping
through
that
pr
for
the
test
service.
F
Yeah
I
had
a
bit
of
an
issue
trying
to
get
the
test
service
and
the
application
talking,
because
I
had
it
set
up
and
docker
composed
in
multiple
containers,
and
I
guess
that
was
just
a
bit
too
complex
and
I
didn't
want
to.
I
spent
the
last
like
two
weeks
trying
to
figure
that
out
and
I
abandoned
that,
and
just
put
it
all
in
the
same
container,
because
I
think
that
approach
was
just
too
complex
and
I
mean
the
whole
point
of
this
is
just
to
run
a
quick
test.
F
So
I've
got
it
working
now.
I
just
need
to
do
a
bit
more
cleanup,
so
I'll
have
that
pr
soon.
F
Yeah
after
that,
there's
a
few
edge
cases
that
I
still
need
to
fix
that
the
the
testing
has
pulled
up
and
then
also
the
the
issue
up
here.
The
span
context
work
that
was
highlighted
in
benjamin's
document.
I
think
four
of
those
bullet
points
are
basically
done
and
there's
one
left
to
the
trace
id
inspanity
export
binary.
C
C
C
C
F
C
Yeah,
I
said
well
like
it's
essentially
just
going
to
be
a
search
and
replace
for
spam
context
to
baggage
like
it's
not
anything
rocket.
Sciencey
is
just
annoying
that
yeah,
I
just
haven't
changed
it
because
I
know
they're
going
to
change
it
or
they're
going
to
try
and
change
it
back,
because
that's
just
how
it's
been
going.
C
F
Yeah
I
can,
I
can
take
that
on
because
I
mean
it's
basically
the
area
I've
been
working
in,
so
I
can
get
to
that
after
these
two
issues
that
I
have
are
done.
I
will.
C
I'll
make
a
ticket
and
I'll
sign
it
to
you.
How
about
that.
C
All
right,
let's
see,
okay,
does
anybody
else
have
anything
else,
I'd
like
to
bring
up
today,
otherwise
I'll
give
everybody
time
back
to
start
working
on
the
pr's,
and
I
got
to
review
a
bunch
of
them.
I.