►
From YouTube: 2021-02-17 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).
D
C
F
C
Yeah
yeah:
let's
wait
for
them.
I
I
think
at
microsoft
you
have
that
this
tradition
of
wait
until
kind
of
all
five
for
any
meeting
to
start.
F
I
remember
that
yeah.
I
wonder
if
that's
still
a
thing
now
with
you
know
instant
zoom
meetings
or
teams
meetings,
I
guess
is
probably
what
they
use.
F
Oh
paulo
did
the
calendar,
invite
thing
or
the
calendar
event
get
fixed.
I
haven't
added
it
yet.
C
Since
it's
fixed,
they
they
they,
I
don't
know
if
they
extend
it.
I
think
they
extend
it.
You
know,
but.
F
C
A
C
F
C
E
E
Yeah,
all
of
the
docs
for
getters
said
not
to
use
any
of
the
native
applications
and
to
just
use
the
web
interface.
C
All
right,
I
think
I
think
it's
a
good
time
for
us
to
start.
I
think
first
thing
I
would
like
to
give
robert
and
david
a
chance
to
introduce
themselves
present.
They
are
just
joining
recently
splunk,
but
they
can
present
themselves
and
tell
about
them
and
they
are
going
to
be
working
really
a
lot
with
us.
B
Okay,
so
let
me
start
so
hi,
I'm
robert,
so
basically
I'm
from
krakow,
david
as
well.
I
don't
know
what
you
want
to
know,
maybe
personal
things,
because
often
that
matters
most
I'm
a
gay,
basically
who
sometimes
speaks
too
much
and
but
but
at
the
same
time
I'm
if
I'm
complaining
or
something
it's
not
about
just
complaining
by
another
positive
person.
So
if
I'm
just
asking
questions,
if
something
is
not
working,
it's
not
just
to
say:
oh
it's
very
wrong.
Sometimes
I
just
want
to
fix
stuff
and
understand.
B
What's
how
how
stuff
is
working
just
to
be
more
productive
and
you
know
spin
up
very
fast
to
the
project
and
maybe
a
little
background
from
technical
perspective.
Last
two
years
I
have
been
developing
golang.
I
have
never
been
using
for
example.net
core.
This
is
my
first
time,
but
before
I
was
using
dotnet
framework
pretty
extensively,
I
was
even
reading.
Writing
an
antenna
ramp
somewhere
using
windows
event
tracing
like
three
years
ago,
or
something
like
that.
So
I
pretty
understand
the
internals
of
both.net
and
windows
pretty
well
and
yeah.
D
Yeah,
so
I'm
david,
I
don't
know
what
actually
you
want
to
know
as
well.
So
yeah,
I
probably
will
be
the
quiet
one,
because
I
prefer
to
listen
what
you
have
to
say
and
just
give
my
five
cents.
If
I
have
something
well,
I've
been
working
with
dotnet
for
a
few
years
now,
no
and
hi
guys.
H
F
C
F
Yeah
I
was
thinking
I
could
introduce
myself
to
since
I'm
immediate
to
you,
robert
and
david
wait
david.
It.
F
Oh,
I
I
one
of
my
teammates.
He
goes
by
pavel,
even
though
it's
like
spelled
with
a
w.
So
I
didn't
really
know
you
know
pronunciation
for
some
people's
names
say
I'm
zach
montoya.
I
work
at
datadog,
one
of
the
original
people
that
was
working
on
the
tracer.
Both
me
and
greg
are
gonna,
be
pretty
active
or
like
the
most
active
from
our
side
at
daviddog,
and
so
I'm
already
really
appreciative
of
the
work
that
you've
already
opened
up
new
pr's
and
just
like
hey.
F
This
looks
wrong,
or
this
looks
weird.
Let
me
fix
that.
So
I
will,
you
know,
try
to
be
responsive
to
pr's
and
stuff
as
much
as
I
can,
but
yeah
I've
been
here
for
a
couple
years
and
have
a
pretty
good
knowledge
of
just
like
our
tracer.
So
if
you
have
questions
about
how
the
current
project
is
and
how
some
of
it
works,
I
can
be
a
resource
there.
Yeah.
E
I've
been
at
new
relic
for
three
years
now,
working
on
the
dotnet
instrumentation
prior
to
that
I
I've
worked
in
the.net
space
for
over
10
years,
working
on
web
apps
and
other
server-side
applications,
and
so
I've
got
to
play
around
with
a
lot
of
different
technologies
and
work
on
some
homegrown
monitoring
stuff
as
well.
I
Hey
guys,
sorry
for
being
late,
I
see
we
have
new
people
super
excited
about
this.
It's
like
two
folks
from
from
splunk
right,
robert
and
david.
E
I
Nice
welcome
guys
very,
very
excited.
I'm
greg
also
at
datadog
doing
all
sorts
of
stuff
with
net
tracing
profiling
and
open
telemetry
kind
of
things
and
yeah
you're
right.
Zac
and
myself
are
probably
the
most
active
in
this
community
here
from
datadog,
but
there
are
some
other
folks
from
our
side
too.
I
Previously
I
worked
at
microsoft.
Doing
monitoring
from
that
side,
so
I
have
a
whole
lot
of
perspective
from
on.net
and
on
how
microsoft
tends
to
monitor
things
and
yeah
really
excited
to
continue
working
here.
I'm
glad
that
we're
growing.
So
I'm
glad
that
we
are
about
to
make
some
more
progress,
cool
and
I
don't
know
whether
did
I
miss
the
introductions
of
david
and
and
and
robert.
H
B
I
Yeah
a
lot:
actually,
where
are
you
located.
B
H
I
Okay,
yes,
we
do
have
about
a
little
less
than
half
the
people
for
our
team
in
various
european
time
zones.
F
Yeah,
if
you
see
tony
tony
redondo
on
some
of
the
github
stuff,
he
works
from
madrid,
so
pretty
close
to
that
time.
Zone.
I
So
if,
if
you
have,
if
we
have
a
significant
proportion
of
people
who
are
in
europe
who
would
like
to
like,
I
would,
I
would
suggest,
if
we,
if
you're,
only
taking
this
meeting
once
in
a
real
while,
then
maybe
we
can
kind
of
really
appreciate
your
flexibility
in
your
time.
But
if
you're
taking
part
regularly,
then
what,
for
example,
the
sdk
sig
group
does?
I
Is
they
alternate
times
that,
like
one
week,
is
more
focused
on
pacific
time
zone
and
one
week
is
earlier
so
that
the
european
people
have
an
easier
time.
C
Yeah,
I
think
I
think
this
is
something
that
we
should
look
in
down
the
road
to
to
have
this
flexibility.
So
people
in
eastern
europe
have
a
better
time
to
join
us
in
the
meetings.
C
I
I
I
would
like
to
start
with
the
thing
that
so
far
I
just
heard
good
things
for
us
to
move
to
is
lack
if
there
is
no.
If
there
is
no
one
kind
of
wanting
to
keep
in
guitar,
I
I'll
do
this
later
this
week,
I'll
create
the
the
channel
and
I'll
post
in
the
guitar.
The
only
thing
is
that
the
the
the
governance
board
had
been
asking
kind
of
for
the
sikhs
to
choose
one
or
other.
You
know
so
kind
of
when
we
go
to
his
black
is
really
slack.
C
I
Just
one
question:
for
slack:
there
is
a
web
interface
as
well.
I
think
there
is
because
I
only
use
it
for
the
client,
but
there
is
also
a
web
interface
right.
C
I
And
do
you
know
how
much
because
one
thing
that
I
like
about
github
it's,
it
goes
relatively
far
in
back
of
history
and
slack
has
limitations,
but
I'm
not
sure
what
the
depends.
B
I
B
I
For
forever,
as
long
as
like,
if,
if
so
that
means,
if
you
post
not
much,
then
it
will
stay
in
there
forever.
B
I
mean
it's
like
the
history
can
be
searched
in
ten
thousand
messages.
This
is
the
promise,
probably
it's
a
little
more
than
that
one
than
that,
but
if
you
know
the
time
nothing's
here
about
time,
I
cannot
find
anything
regarding
time.
I
Yeah,
I
would
sort
of
look
into
this
because
this
can
turn
out
to
be
a
a
caveat.
My
main
concern
with
slag
is
that,
if
essentially,
I
really
would
don't
want
to
be
in
a
situation
where,
if
I
haven't
looked
at
the
message
within
a
few
days
after
it
was
sent-
and
I
entirely
missed
it-
so
that's
the
situation
that
I
would
like
to
avoid
other
than
that.
I
have
no
concerns.
C
I
There
is
apollo,
there
is
also
there
is
definitely
some
kind
of
time
limit.
I
know
this
from
internal
data
dock
usage
that
if
you
put
something
into
slack
and
then
you
go
to
look
at
that
message
after
several
weeks-
it's
no
longer
there.
I
don't
know
what
exactly
this
limit
is,
and
if
it's
really
many
weeks,
if
it's
like
over
a
month,
then
it's
not
a
problem.
Yeah.
C
C
But
I
I
I
would
like
us
to
have
kind
of
or
reach
a
decision
now,
if
you're
gonna
try,
because
also
if
we
try
and
there
is
any
problem
we
can
switch
back.
I
I
would
like
to
avoid
that,
but
I
think
we
should
kind
of
give
a
try,
and
that
is
the
problem.
We
switch
back.
I
If
it's
like
a
month,
I
have
no
concerns.
I
just
want
to
make
sure
it's
not
like
you
know
you
didn't
don't
check
for
a
few
days
and
suddenly
it
disappeared
because
we
use
some
free
tier
or
something
like
that.
So
that's
that's
that
that
would
be,
I
think,
a
deal
breaker
for
sure.
C
Yeah,
so
I
let's
do
the
follower
we
keep
the
the
guitar.
I
I
will
be
I'll
put
a
reminder
for
me
to
check
kind
of
every
three
or
four
days,
just
in
case.
B
I
just
checked-
I
just
checked
with
the
slack,
so
I
have
to
read
too
fast.
It's
it's
written
that
that
the
starter
plant
can
search
beyond
10
000
message
limit
and
there
is
no.
So
it's
just
like
comparison
to
free
plan
and
if
I
see
the
time
comparison
that
we
are
there's
standard,
this
unlimited
message
are
here
archive.
C
Okay,
so
let
let's
be
doing,
I
will
do
the
switch.
I
will
send
a
message
to
the
guitar
channel
and,
as
I
said,
that
will
be
very
fine
kind
of
every
two
or
three
days.
If
somebody
post
and
if
somebody
post
there,
I
will
direct
them
to
the
slack
and
if
down
the
road,
we
find
a
problem
we
switch
back
to
together
all
right,
I
would
like
to
kind
of
the
second
topic.
I
would
like
to
follow
up
on
the
question
about
the
expand
data
model.
C
That
was
that
issue
that
we
had
some
discussion
via
github,
not
twitter,
github,
and
if
there
is
anything
that
they
see,
can
do
to
help
on
that.
I
think
we
we
start
to
have
some
conversation,
and
perhaps
robert
wanted
to
start
to
do
some
work
on
that.
But
I
would
like
to
to
check
with
greg
and
zach
kind
of
where
we
are
kind
of
we
decide,
or
you
would
like
to
see
a
proposal
in
code
for
that.
I
Kevin
today,
who
is
the
performance
owner
on
our
team
and
because
he
didn't
provide
any
feedback
on
this,
and
I
thanked
him
today
last
like
please,
please
do
because,
like
we
chatted,
it
seems
to
like
zack
from
your
perspective.
Does
it
generally
look
fine.
F
Yeah
I
was
proposing
as
well
like
something
an
optional
decorator
for
the
on
method,
end
in
case
like
status
code
stuff.
That
requires
the
response.
That's
basically,
it.
I
Yes,
I
think
this
is,
I
agree,
but,
like
it's
it's
details
of
how
to
make
this
better.
It's
not
a
jailbreaking.
I
I
I
don't
have
objections
from
kevin
and
here
for
him
it
got
buried,
and
I
asked
him
today
to
kind
of
get
this
going,
but
it
looks
like
we
can
essentially
start
coding
on
this.
The
idea
would
be
like.
Ideally,
if
we
wait
for
two
three
days
and
if,
unless
unless
we
are
free
and
we
are
blocked
by
this,
then
we
can
start
if
we
are
doing
other
things
that
are
useful.
I
Then
we
can
give
kevin
a
few
days
to
take
a
look
at
this,
so
that
we're
not
by
any
surprises
from
his
feedback,
and
the
idea
would
be
that
somebody
from
the
sig
implements
it
and
then
people
on
the
from
datadog
review
this,
because
once
the
code
is
pushed
into
this
open,
telemetry
master
branch,
we
will
want
to
integrate
it
back
into
data.
C
So
so
I
think
you,
as
you
said
you
you've
been
kevin
and
one
of
us
or
I
I
think
likely
robert
will
start
to
kind
of
explore
the
implementation
of
that.
I
think
he
has
a
few
things
before
that,
but
then
perhaps
we
already
have
heard
back
from
kevin
by
that
time
and
he.
C
All
right,
so
the
next
thing
we
have
with
robert
and
david
joined
us.
We
have
a
bunch
of
prs,
hoping
I
don't
know
if,
if
you
guys
wanna,
because
some
of
them
having
questions
like,
for
instance,
I
think
this
one
has
the
question
kind
of.
Do
you
really
want
to
work
on
mac
because,
like
the
prepare
tools
version
right
now,
just
building
dotnet
framework
is
a
small
thing,
but
since
david
was
using
the
mac
for
him,
it
becomes
a
thing
that
doesn't
build.
C
You
know,
so
I
think
I
I
think
from
in
general
the
perspective
of
open
source
projects.
Yes,
we
should
be
looking
to
make
it
work
on
every
development
platform.
We
may
not
be
able
to
do
that
in
an
snap,
but
we
should
be
progressing
that
direction.
F
Yeah
agreed
we
these
tools,
which
you
know,
aren't
really
product
code.
We
can
target
those
for
redundant
core
and
case
by
case.
If,
for
some
reason
like
we
need
a
framework,
we
can
do
that,
but
certainly
we
can
try
and
make
these
tools
work
just
as
dynamic
rats.
H
C
Yeah
yeah,
we
we
use
jet
brain
tools
also
to
be
fair.
My
my
development
environment
nowadays
is
visual
studio,
visual
studio
code
jetbrains.
You
know
because
each
one
of
them
does
one
thing
better
than
the
other,
so
I
keep
jumping
between
them.
I
Well,
the
back
end
of
this
is
tony
works
on
the
mac
and
he
was
like.
I
want
this
to
work
on
the
mac,
and
here
just
you
know
on
on
some
friday.
It
wasn't
working
on
a
mac
and
then
on
monday,
after
that
friday
it
was
it
was,
and
everybody
went.
Oh,
that's
cool,
so
that's
the
background
of
it.
I
I
don't
know
I
I
don't
mind.
E
F
Yeah
I'll
try
to
I'll
try
to
take
a
look.
I
just
got
back
from
pto,
so
I
just
missed
most
of
these.
C
Yeah
yeah
and
chris
here
really
asked
good
questions
I
after
he,
he
mentioned
like,
for
instance,
this
sample.net
framework
to
zero.
Okay,
do
you
really
want
to
make
it
work
on
core
or
just
one
skip?
You
know.
D
B
Think
the
only
solution
for
this
one
is
make
a
separate
solution
fire
to
be
honest,
because
I
was
trying
to
read-
and
it's
not
possible
to
ins,
or
maybe
I
found
something,
because
if
there
are
some
projects
which
are
targeting
like
your
only.net
framework,
we
need
to
make
sure
that
they're
not
loading
right.
Otherwise
it
will
be
problematic
and
they
have
the
same
problem
with
vs
code
with
with
omnisharp
or
how
was
it
called.
C
Yeah,
so
we
need
to
to
come
to
our
question.
One
possibility
is
that
it's
kevin
or
tony
that's
using
the
mac
drag.
C
Okay,
because
I
think
visual
studio
for
mac
does
install
a
lot
of
targeting
packages
that
are
kind
of
visual
studio
stuff
and
that
allows
to
build
a
bunch
of
stuff
that,
typically,
if
you
just
install.net
without
the
visual
studio
for
mac,
you
can't
build.
I'm
I'm
not
saying
I
didn't
try
this.
I
don't
have
a
mac
anymore,
so
I
don't
know,
but
there
was
this
in
the
past
when
I
tried
visual
studio
for
mac.
C
I
think
this
was
one
of
the
the
issues
the
other
one.
This
is
was
a
thing
that
I
did
in
the
past,
and
this
ties
with
another
thing
that
I
would
like
to
talk
in
this
meeting
today
about
the
the
stuff
with
the
active
source.
C
Let's
talk
later,
but
I
had
added
this
these
things
to
the
master
in
the
past
long
time
ago,
but
I
think,
since
we
have
that
branch
that
greg
worked
in
the
la
in
the
end
of
last
year,
I'm
thinking
in
removing
this
from
master,
because
we
are
not
really
doing
that
to
work
on
master
on
the
master
branch.
So
I
think
we
should
remove
the
open
telemetry
and
the
vendor
stuff
from
diagnostic
source
from
the
master.
C
C
I
But
hold
on
because
once
we
restart
work
there,
we
will
be
merging
master
into
the
future
branch.
Yes,
if
you
remove
this
from
master
now,
that
means
a
deletion.
That
means
when
we
merge
most
into
the
feature
branch,
it
will
remove
the
stuff
from
the
future
branch.
C
I
Basically
sure
if
we
validate
this,
like
I
I
agree
with
like
not
let's
not
have
stuff
that
is
only
necessary
for
a
feature
in
in
master,
that's
cool,
I
just
what
I
don't
wanna
do.
Is
I
basically,
when
I
merge
from
mastering
to
feature
branch,
I
want
this
to
be
a
conflict-less
merge.
I
If
that's
given,
then
no
worries.
So
maybe
if
we
make
this
removing
then
try
to
merge
from
there
into
the
feature
branch
and
if
it
works,
then
the
very
good
should
go.
But
if
it
doesn't,
then
maybe
we
should
look
for
another
way
of
getting
there.
C
Okay,
I'm
gonna
validate
this
location
and
validate
the
merge
before
we
actually
do
any
removal
just
to
be
safe,
all
right,
and
then
I
think
there
are
the
things
are
other
prs.
Do
you
have
pending
questions
that
anyone
would
like
to
discuss?
C
I
think
robert
proposed
something
getting
the
sdk
instead
of
the
runtime,
which
saves
some
space.
This
is
about
a
comment.
I
still
didn't
look
to
the
pia,
but
this
is
about
a
comment
from
when
I
introduced
the
zip
king
exporter,
in
that
we
talked
about
instead
of
passing
the
settings
passing
the
source
of
the
config,
so
the
exporter
could
have
its
own
reader
of
the
configuration
without
touching
the
the
settings.
So
I
I
think
that
is
the
the
original
intent
I
still
rather.
B
B
C
Okay,
I
still
have
to
look,
but
if
you
have
any
questions
for
for
the
people
here
then,
but
it
seems
that
this
is
is
good
and
then
you
had
some
ideas
about
integration
tests
right.
You.
C
B
Skip
it,
maybe
maybe
if
we
had
time
at
the
end,
I
can
just
quickly
explain
my
thoughts,
but
I
think
that
the
second
pair
from
the
bottom
is
the
most
important,
because
all
the
time
there
are
some
unit
tests
that
are
failing
and
personally
prefer
to
have
the
the
continuous
integration
green
as
much
as
possible.
Even
for
us
who
are
new
to
be
confident
that
you
know
whenever
we
do
something,
it's
really
working
not
just
and
re-run
and
waste
two
hours,
because
it's
not
really
working.
D
C
Is
basically
remove
the
dd
and
stuff
we
we
tried,
you
tried
to
kind
of
make
it
conditional,
but
there
was
no
condition
in
the
azure
configuration
files
to
just
skip
the
test.
So
in
the
end
you
have
basically
to
remove
dd
agent
related
right,
correct,
okay
and
just
to
not
miss
so
the
video
set,
the
the
one
on
the
bottom
right.
C
Oh,
okay,
yeah!
This
is
very
simple,
but
I
just
want
a
confirmation,
because
I
think
this
is
showing
up
on
your
build
on
mac
right
of
it.
B
F
C
So
the
other
thing
that
I
want
to
talk
about
is
because
we
are
doing
this
push
in
this
direction.
That
gets
us
to
have.
C
Telemetry
but
I
I
would
like
to
not
forget
the
activity
and
active
source
path
that
we
can
develop
and
you
should
develop
eventually,
but
I
I
I
would
say
that
initially
from
our
side,
we
are
going
to
be
prioritizing
the
other
stuff.
But
I
want
to
be
sure
that
we
do
some
work
with
activity
and
activity
source
that
greg
had
already
started.
Oh
and
he
he
needed
to
to
go
right
now.
But
I
I.
C
Yeah,
so
I
I
just
want
to
be
sure
that
we
have
some
kind
of,
even
if
it's
not
immediate
right
now,
that
this
kind
of
doesn't
follow
off
our
of
the.
J
C
And
kind
of
is
still
we
make
clear
for
people
from
outside-
that's
still
part
of
the
plan
and
that
we
are
going
to
do
work
in
that
direction.
You
know.
I
I
actually
have
some
hopeful
news
about
this,
so
first
of
all,
it's
it's
basically
my
fault.
I
I
kept
meaning
to
get
back
to
this
and
finish
this
up,
and
I
kept
promising
this
and
there
was
always
some
other
organizational
priority
that
was
assigned
to
me.
I
So
I
kind
of
slipped
it
for
like
three
months
now,
but
we
have
a
kind
of
prototyping
week
coming
up
next
week,
where
we
work
on
sort
of
product
related
but
non-project
not
not
scheduled
things,
and
I
am
currently
in
like
90
sure
that
I
can
work
on
something
related
to
open
telemetry
and
what
I
would
like
to
do,
because
it
needs
to
be
scoped
for
a
week.
It's
like
after
the
week
is
finished.
Then
I
have
to
stop
doing
and
switch
back
to
my
normal
tasks.
I
I
I'm
thinking
and
please,
if
you
have
feedback
about
what
I
can
do
otherwise,
then
I
would
love
that,
but
I'm
thinking
taking
what
we
have
for
versioning
and
the
problems
around
versioning
and
scoping
it
down
to
solving
the
problem
for
diagnostic
source.
Only
so
that
would
mean
that
we
have
to.
I
We
have
to
that
would
include
solve
the
library
loading
correctly
and
rendering
and
all
the
smarts
around
that
and
then
like.
We
have
this
unfinished
activity.
Stub
api,
ignore
that
for
a
moment,
but
create
the
same
style
diagnostic
source,
tab,
api
that
uses
the
same
code
emit
logic
to
make
it
fast
and
essentially
creates
an
api
that
allows
you
to
register
listeners
for
diagnostic
source
via
the
stub
without
taking
dependencies
and
processing
these
events,
and
that
would
be
like
a
library.
I
The
the
outcome
would
be
that
in
the
tracer,
wherever
we
are
currently
using
diagnostic
source
as
a
source
of
telemetry,
we
can
start
using
that
that
thing
that
I'm
proposing
that
I
built
during.
C
I
Getting
this
finished,
we
can
include
this
into
the
product
both
on
the
side
of
data
dog
and
on
the
site
of
open
telemetry,
and
that
will
give
us
hardening
in
terms
of
knowing
that
the
versioning
loading
stuff,
because
that's
the
same
for
activities
right
that
that
is
done
well
and
then,
as
a
subsequent
effort.
In
some
future,
we
can
go
back
and
do
and
continue
the
stuff
that
we're
doing
for
activities.
C
I
Less
apis,
just
yeah
on
on
on
on
activity
source,
there
is
like
many
like,
maybe
five
times
less
or
maybe
ten
times
less
on
activity
so
as
compared
to
certain
diagnostics
compared
to
activities
and
the
whole
additional
logic
related
to
making
activity.
Ids
work
sensibly
also
does
not
exist
on
diagnostic
source.
So
that's
why
it's
way
less
work
and
can
be
scoped
way
down.
C
C
H
C
I
So
so
because
I
need
to
to
make
sure
that
I'm
done
in
a
week
it's
something
that
is
useful
to
everybody,
not
just
like
a
thing
that
is
almost
done,
but
nobody
knows
how
it
works,
and
so,
let's
use
this
right.
I
will
probably
not
attempt
to
make
this
extendable
in
terms
of
this
proposal
that
we
said
like,
so
that
different
different
tags
can
be
plugged
in.
I
I
will
just
make
it
work
with
what's
already
in
the
tracing,
so,
whereas
the
sig
can
help
is
taking
this
idea
that
I
outlined
in
this
issue
for
making.
I
Io
based
instrumentations
extendable
for
different
tags
and
apply
essentially
the
same
thinking
some
sort
of
easy
way
to
to
to
that
new
code
so
that
we
can
start,
you
know
start
using
it
so
that
something
where
I
think
I
would
I
would
appreciate
something,
and
that
week
is
that
week
starts
next
week's
in
the
middle
of
next
week.
So
that
means
in
a
week
and
a
half
from
now.
This
is
already
going
to
be
done.
C
Okay,
so
I
I
think
that
then
we
we
wait
on
on
your
own
work
with
the
diagnostic
source.
So
then,
when
that
we
have
that
on
the
branch,
then
we
can
start
try
to
to
do
the
stuff
to
really
deliver
that
as
an
instrumentation,
with
the
open,
telemetry,
semantics.
C
Yes,
yes,
okay,
okay,
yeah
see
seems
seems
like
this
is
is
is,
is
it
still
progressing
and
yeah
yeah?
I
want
to
keep
kind
of.
As
I
said,
I
want
to
keep
asking
about
this,
so
we
kind
of
ensure
that
it
doesn't
get
forget
for
god.
C
F
F
Our
ci,
since
we've
identified
there's
at
least
like
10,
maybe
15
tests
that
are
flaky
and
so
with
so
much
stuff
going
on
we're
trying
to
try
to
improve
that,
and
so
I
do
want
to
know
that
there
are
some
flaky
tests
and
we
will
try
to.
You
know
constantly
improve
on
those.
But
if
you
see
some
stuff,
please
reach
out
and
let
us
know
like
hey:
does
this
look
expected
or
not?
F
We
have
a
couple
of
unit
tests,
mostly
it's
just
integration
tests
where
maybe
sometimes
it
doesn't
shut
down
cleanly
or
something
so
just
gonna
bring
that
up,
especially
for
new
contributors
that
just
to
know
that
there
are
some
issues
we're
still
trying
to
work
through.
But
that's
not
it's
not
unexpected.
B
Basically,
maybe
I
do
not
want
right
now
to
go
into
code
because
I
just
want
to
first
talk
about
high
on
a
higher
level.
B
Basically,
first
thought
thing
that
stuck
me
was
that
I
don't
know
if
it's
good
or
not,
that
there
is
this
docker
compose,
which
spin
ups,
everything
you
know
like
ready,
select,
t-shirt
five,
five
six
and
I
just
wondering
if
it's
not
something
that
makes.
I
don't
know
some
time
outs
on
this
on
the
you
know,
on
the
machines
etc
or
it's
not
that
that
they
are
flaky.
You
know.
Sometimes
I
haven't.
F
Seen
I
haven't
seen
the
docker
compose
part
be
an
issue
there.
I've
seen
we
have
seen
timeouts
and
try
to
identify
some
where
perhaps
like
a
like
an
s
unit
test
might
hang
or
something,
but
from
the
docker
compose
side,
starting
up
the
services
and
hasn't.
I.
B
B
C
B
F
Always
doing
compose
so
the
usual
dev
flow
for
me
is
I
we
have
so
for
one
reason
that
it
takes
so
long.
Is
we
test
like
a
matrix
of
versions,
and
so
there's
like
the
one?
That's
nci
it'll
test,
a
larger
matrix
of
like
package
versions
like
for
different,
like
mongodb
clients,
we
do
less
for
the
local
testing,
and
so
that's
still
kind
of
in
the
details
and
complex.
We
do
have
some
docker-composed
ones
where
it
will
just
run
that
particular
service
or
that
particular
app.
F
So
if
you
wanted
to
test
just
like
the
mongodb
one
like
it
doesn't
run
through
the
whole
infrastructure,
it
would
just
be
like
here's
set
up
the
and
then
run
the
sample
application
that
corresponds
to
it.
But
it
doesn't
go
through
that
integration,
test
infrastructure
like
you're
saying,
but
mostly
I
run
integration
tests
through
ci
and
I
just
test
like
one
stamp
application
at
a
time
as
needed
locally
for
my
on
my
windows,
machine
yeah.
F
We
don't
have
a
good,
isolated
way
to
do
like
one
test
at
a
time
for
I
guess
like
for
mac,
we
have
the
ability
to
there's
a
docker
compose
that
starts
up
and
then
runs
each
all
the
integers
just
at
once,
but
that's
it
right
now.
F
Yeah
yeah,
I'm
actually
I'm
not
even
familiar
with
what
the
build
process
looks
exactly
like
for
mac,
because.
I
Mac
is
we
made
it
work
and
it
does
work,
but
it's
not
an
official
product.
I
So
if
suddenly
we
come
to
a
conclusion
that
it's
more
trouble
than
it's
worth,
we
may
decide
drop
it
yeah,
but.
I
H
I
For
now
it's
so
like,
I
definitely
wouldn't
want
to
like
go
either
way
without
chatting
to
tony,
because
he
seems
to
to
make
it
work
and
like
if,
if,
if
there
are
good
arguments
to
remove
it
from
both
datadog
and
the
open
telemetry
only
from
open
climate
or
whatever,
like
generally
yeah,
it's
not
an
official
official
support.
But
if
it
works
with
a
little
effort.
Why
not.
F
I
do
think
I
I
do
think
the
the
route
you're
pursuing
about
setting
up
a
picture
should
work
pretty
well,
because,
when
we're
running
the
integration
tests,
we
it
there's
one
test
collection
that
runs
each
like
all
the
different
package
versions.
So
if
you
have
the
one
fixture
for
that
collection,
then
that
works
pretty
well
and
there's
there's
a
direct
mapping
of
like
mongodb
there's
one
a
sample
application
that
uses
it
and
then
for,
like
my
sequel,
there's
one
that
uses
that.
So
it
does
work
out
really
well
that
way.
F
F
H
F
Did
you
want
to?
Did
you
want
to
set
up
like
more
time
to
talk
through
it
a
little
more
outside
the
sig?
Would
that
help
or
I
don't
I
don't
want
you
to
get
like
stuck
on.
B
F
B
I
You
look
at
the
issue
about
abstraction.
He
has
a
commander,
then.
E
I
B
F
Yeah
yeah,
sorry,
I
was
nodding,
but
we
got
friends
in
the
back
making
a
call.
So
I
muted
myself.
F
I
Just
want
to
repeat
how
excited
I
am
that
more
people
are
joining.