►
From YouTube: 2021-06-10 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).
B
C
C
Yeah
I
saw
a
couple
of
them.
I
watched
the
sort
of
the
deep
dive
on
the
general
changes
that
are
coming
like
the
xcode
cloud
and
the
swift
async
support,
which
is
pretty
cool.
B
Yes,
yeah.
I
follow
all
the
async
await
sessions,
the
also
the
actors
one
and
the
actors
yeah
all
that
stuff,
mainly
mainly
those
yeah,
because
I
thought
it
was
the
most
interesting.
A
B
That's
something
that
we
will
have
to
handle
for
sure
yeah.
C
B
Yeah,
in
fact,
I
put
some
some
lines
in
the
in
the
topics
for
today
about
that.
B
It's
a
little
opinion
that
yeah
mainly
that
I
think
that
right
now
they
are
only
supporting
the
still
to
come
versions
for
ios
and
mac.
So.
B
Not
that
I
at
least
not
until
today
and
yesterday,
but
I
I
I
don't
think
they
have.
They
had
also
that
sweet,
lock
library
and
never
talked
about
that
in
the
last
three
years.
So.
A
B
B
I
don't
know
why
but
yeah.
Oh,
we
need
this
here.
D
I
saw
your
comments
from
last
week.
B
We
were
talking
about
doing
about
interesting,
instrumentation
things
we
could
do,
but
we
didn't
take
any
decision
about
that.
Yet
talked
about
possibilities
and
the
near
future,
but
yeah,
nothing,
nothing!
B
Okay,
here
we
are
okay,
for
me,
has
been
a
quite
busy
week,
work
at
my
company,
so.
B
We
are
putting
in
a
private
beta,
the
product
that
we're
developing
with
open,
telemetry,
so
yeah
it
was
quite
busy,
but
let's
see
what
we
have
for
reviewing
last
last
meeting
actions.
Okay,
sweet
packages
to
the
registry
still
need
to
be
done.
Okay,
I
will
copy
this
exactly
like
it
is
the
open
telemetry
registry.
You
know
we
should
have
our
library
there
just
to
be
known
and
the
things
that
we
instrument
that
also
can
be
referenced
at
the.
B
If
someone
search
for
that
yeah,
I
would
yeah,
I
said
I
would
try
to
find
out
time,
but
I
couldn't
so.
I
will
yeah.
I
will
do
it
probably
tomorrow
morning
as
soon
as
possible,
because
it's,
I
just
thought
it
would
take
too
long.
B
Okay
for
the
code
coverage
yeah,
I
have
asked
about
the
codecov
account,
but
I
have
done
it
today
when
I
was
reviewing
the
new
things
to
be
done.
So
I
still.
B
We
have
some
kind
of
account.
I
have
never
used
code
code,
but
I
look
like
it
needs
an
account.
I
don't
know
how
it
works
with
with
registering,
and
so
I
asked
the
more
mobile
career
info
was
closed,
so
that
no
things
do
there.
C
I
was
trying
to
get
a
hold
of
carlos
and
I
I
think
he
might
just
be
really
busy.
I
asked
him
last
friday
if
we
could
get
together
this
week
sometime
to
chat
about
like
what
needed
to
be
done
for
that,
because
it
was
kind
of
unclear
that
it
seemed
like
there
was
open
issues
that
I
hadn't
addressed,
but
I
didn't
really
feel
that
way,
but
yeah
I
just
can't
get
hold
of
them,
so
maybe
I'll
just
try
reopening
it
and
seeing
where
it
goes.
C
Yeah
yeah,
I
think
I
think
that's
a
good
idea,
because
that'll
help
remind
me
to
to
to
keep
it
moving.
B
Yeah
that
yeah,
I
think
it's
a
good.
I
would
way
yes
to
to
keep
your
things
yeah
to
keep
attention.
B
Okay,
the
my
internet
release
was
released.
I
was
thinking
about
doing
another
maintenance
release
with
the
latest
changes.
I
asked
james
to
test
your
pr.
I
don't
know
if
he
has
said
that
he
still
hadn't
yes
yeah
but
yeah.
We
we
also
had.
I
don't
know.
I
think
we
had
some
other
some
other
things
fixed.
Be
not.
Could
you
test
the
jager
exporter
time?
I
I
didn't.
D
Okay,
yeah,
I
will.
I
will
test
it
out.
B
Okay,
maybe
I
I
think
I
I
will
release
another
version
with
your
peers
about
so.
B
B
Maintenance,
okay,
more
your
peer
was
merged,
but
still
james
didn't
respond
so
and
you
are
now
an
approver
also-
and
we
talked
about
libraries
to
instrument,
but
we
didn't
take
anything
about
that.
B
I
think
it's
a
good
idea,
I
think
the
network,
the
the
james
idea
is
already
you
know,
is.
B
An
issue
but
yeah,
I
think,
would
you
mind
creating
issues
for
that
in
the
project
yeah.
B
B
Yeah,
so
yes,
people
can
take
ideas.
Also,
maybe
someone
wants
to
add
some
er
there,
sometimes
from
time
to
time.
Someone
appears
and
and
creates
a
pr
for
the
project.
Yeah.
C
B
I
don't
know
how
to
get
that
people
to
participate
in
a
more
constant
way.
C
It's
really
maybe
we
need
to
make
t-shirts
or
something.
C
It's
like,
you
know,
make
10
pr's
and
you
get
a
t-shirt.
B
Okay,
so
I
think
that
that's
all
yeah,
if
you
can
create
those
issues.
B
B
Okay,
the
next
you
know
you
have
any
other
topic
just
to
add
yeah
for
one
of
the
topics
I
I
wanted
to
bring
here
is
the
the
new
async
await
technology.
That
apollo
is
bringing
to
swift
5.5
for
the
next
release
of
a
code.
B
Yeah
it
from
what
is
said
in
the
documentation.
Also
answered
by
the
apple
engineers
is
that
it's
currently
only
supported
on
what
will
be
ios,
15
and
mac
os
12..
B
So
we,
I
think
we
should
not
rely
on
that
technology
currently,
because
we
will
end
up
with
incompatible
libraries
users.
There
are
some
really
cool
things
like
the
actors
and
the
async
await
for
some
of
the
issues
we
have,
for
example,
for
exporting,
and
things
like
that
that
we
we
had
the
problem
that
we
we
had
to
wait
for
the
export
to
know
the
result
currently.
C
A
B
D
B
Changed
some
things
in
the
way
threats
are
handled
and
that
stuff
that
makes
sense.
Oh
it
needs
some
work
on
the
operating
system.
Maybe
they
have,
they
will
bring
some
update,
for
maybe
the
current
version.
Yeah
ios
14
will
have
support
for
that,
but
I
don't
expect
it
to
go
much
much
further
than
that,
they
didn't
say
even
ios
14,
but
when
they
have
released
this
kind
of
technology
usually
support
the
newer
and
the
previous
one.
B
C
C
C
A
B
C
Well,
I
saw
that
video
on
on
the
actor,
and
that
was
like
exactly
what
I
just
did
in
that
pr.
A
couple
of
days
ago,
where
I
had
to
like-
and
I
was
like
damn-
I
needed
this.
D
B
Yes,
just
for
free,
yeah,
okay,.
D
So
so
are
there
third
party,
you
know
swift
libraries
that
you
know
are
used
today
for
like
instrumentation?
No,
are
they
popular,
like
you
know,
actor
models
unless
inc
away?
Oh.
D
C
I've
never
really
looked
into
that.
Have
you
seen
any
like
async
await
style
libraries
to
help
with
concurrency
nacho.
B
In
in
swift
now
some
I
mean
something
like
features,
something
like
and
things
like
that
there
were
some
features.
B
B
I
don't
know
what
was
that
it
was
quite
new
library
for
having
something
similar.
Also,
what
was
the
name
something
with
it
was
a
complex
library.
I
never
used
that
because
I
thought
it
was
too
complex,
mostly
that
it
provided
on
top
of
the
usb
the
ecd
quiz
queues.
I
don't
know,
what's
the
name
of
that
library
data
for
this,
I
think
it
was
two
years
ago
that
they
had
that
library.
B
D
It
was
such
important,
so
the
instrumentation,
you
know
that
you
guys
are
talking
about
additional
frameworks,
are
totally
based
on
what
is
supported
by
apple
and
what
is
being
used
by
different
customers.
At
this
point,
based
on
the
popularity
is
where
we
start
looking
into
adding
instrumentation.
B
Yeah,
I
think
that
also
the
one
of
the
things
that
had
the
project
was
the
task
class
and
it
can
keep
task
task
properties.
So
we
could
keep
the
context,
probably
inside
a
task
if
we
could
move
to
task.
As
I
said,
but
yeah
we
have
the
problem
with.
B
We
have
the
problem
of
supporting
previous
versions
of
the
os,
so
I
think
that's
a
main
problem
with
with.
B
And
yeah,
so,
let's
see
how
how
that
changed.
Okay,
also
another
thing
about
this-
is
the
next
point
that
it's
a
sync
awaiting
network
instrumentation,
one
of
the
things
that
they
added
with
a
single
weight.
I
mean
using,
I
think,
with
ourselves
in
the
library.
I
think
it's
dangerous,
because
we
will
be
stopping
the
possibilities
of
our
users
to
deploy
in
previous
versions
of
the
operating
systems.
B
So
they
are
also
adding
that
to
objective
c,
because,
as
you
know,
your
session
is
written
in
objective
c
and
it
just
has
the
the
the
wrapper
for
calling
calling
from
sweet.
B
B
If
not,
we
will
lose
that
network
information
and
I
think
that
many
people
is
going
to
use
that
because
it's
also
much
much
simplified
so
yeah.
B
B
That
that
was
my,
we
must
check
what
new.
B
Yeah,
okay,
also
one
of
the
things
that
also
showed
was
something
they
called
doxy
that
it's
a
doc
computer
compiler.
B
We
were
talking
about
this
documentation
thing
some
meetings
ago
about
what
what
were
what
were
the
ways
of
getting
documentation
from
the
source
code.
But
this
doxy
is
exactly
that.
So
if
you
open
the
project
directly
in
doxy,
it
generates
your
documentation
in
the
xc
archive
format
for
documentation,
so
you
can
integrate
it
with
xcode
and
also,
I
think
it
can
generate
docs
in
html,
ml
and
other
formats.
So
I
think
we
can
even
use
that
previously.
B
And
and
just
put
that
code
inside
the
project,
so
people
can
use
that
so
maybe
yeah
I
I
will.
D
B
Yeah
that
that's
right
with
the
new
xcode
it's
now
in
bit
in
beta,
but
the
news
code
they
announced
on
monday,
it
automatically
generates
documentation
from
the
code.
So
I
I
think,
is
the
fastest
way
for
us
to
provide
documentation
to
our
users
so
trying
to
use
that
automatic
thing
would
be
great.
It
shouldn't
take
long,
I'm
probably
checking
a
pair
of
sessions.
That
apple
has
just
to
know
how
it
works
and
make
it
run.
C
Yeah,
if
I
find
some
time
I'll
I'll,
probably
take
a
look
at
it,
I
haven't
tried
installing
xcode
beta
in
a
while.
Is
it
still
like.
A
B
B
Okay,
about
maintenance
release.
I
was
planning
to
release
a
new
maintenance
release
with
the
fix
that
bryce
did.
I
think
it's
important
enough
to
just
have
a
version
and
it's
almost
free
just
to
create
a
number.
So
people
can
just
take
that
in
a
number
in
a
numbered
way
and
automatically
update
that
so.
C
I
wanted
to
talk
about
adding
adding
a
couple
of
like
tracer
classes
to
be
open
rather
than
public.
I
was
running
into
an
issue
we're
trying
to
solve
this
clock,
skew
issue
at
elastic
and
I
think
the
best
way
to
do
it
is
by
extending
the
open,
telemetry
trace
exporter
and
then
adding
a
export
time
stamp
to
the
trace
exporter.
C
The
problem
is
that
we're
using
like,
I
don't
think
that
there's
a
a
semantic
convention
defined
in
open
telemetry
to
name
it,
so
I
don't
want
to
add
it
into
well
like
we're
using
some
like.
You
know,.
C
Telemetry.Elastic.Exporttimestamp-
and
I
don't
want
to
add
that
to
the
open
telemetry
sdk,
so
I
was
thinking
wrapping
the
otlp
exporter,
but
it
needs
to
be
open
instead
of
public,
so
I
can
extend
it
from
outside
the
package.
B
B
I
think
that
the
telemetry,
the
telemetry
sdk,
should
be
a
final
class.
It
shouldn't
be
narrated
by
by
the
spec
I
mean.
B
Spec
says
that
the
that
should
be
the
the
behavior
of
the
of
the
tracer.
E
So,
oh
it
it's
not
the
it's,
not
the
tracer.
It's
the
exporter.
B
C
Do
we,
but
I
I
just
bring
this
up
because
if
we're
gonna
do
a
1.0.2
release,
I
guess
I
can
verify
if
it
is
okay
in
the
spec
to
do
that.
But
here
let
me
sh.
Let
me
show
you
if
I
can
find
that.
Where
is
it.
C
B
E
B
C
Yeah,
let
me
I'm
I'm
looking
for
my
my
zoom
window.
C
I
think
okay,
so
here's
the
here's,
the
the
the
minor
change,
so
we
have
the
open,
tometry
exporter
and
what
I
want
to
do
is
I
want
to
enter.
You
know
like
extend
the
export,
maybe
there's
a
better
way
to
do
this
than
than
creating
a
subclass,
but
just
extend
the
export
function,
so
that
I
can,
let
me
see
here,
is
it
pushed
up
somewhere?
It
must
be.
E
B
I
have,
I
have
seen
it
in
the
pr
you
just
want.
C
C
Yep
yep
yeah
as
long
as
that's
okay
with
the
the
spec
I
can
do,
I
can
do
other
things
to
make
this
work,
but
this
was
just
the
easiest.
The
smallest
number
of
changes.
C
Yeah,
so
yeah,
so
here
is
here's.
What
my
intention
is
so
I
want
to
I'm
sure
here
is
the
elastic
exporter
here.
So
here
I
am
extending
the
otl.
Can
you
see
my
screen
here?
Is
this
too
too
small?
Okay,
so
extending
it
calling
the
export
function
grabbing
an
export
timestamp
to
add
to
each
of
the
spans
using
this
telemetry.sdk
elastic
export,
timestamp.
C
B
B
B
Yeah
we
should
check
the
the
the
spec
we.
We
could
also
add
some
hooks,
so
we
that
could
be
done
also,
maybe
better
than
in
everything.
Having
hooks
for
modifying
your
but
your
spam.
C
Yeah
yeah,
the
so
the
yeah.
The
key
is
when
these
spans
are
actually
sent
up
to
the
to
the
collector
to
the
yeah
to
the
collector.
It
needs
a
time
stamp
of
when
it
was
sent
to
try
to
clamp
down
the
the
clock
skew
on
the
timestamps
in
of
the
actual
creation
of
the
spans
themselves,
so
that
they
can
try
to
do
like
some.
C
Yeah
yeah,
it
will
still
be
some
degree
of
time
difference,
but
it
will
be
able
to
kind
of
sort
it
out
a
little
bit.
B
I
don't
I
don't
see
a
problem
with
this.
We
we
can
check
if
that's
possible.
If
not,
we
could
also
add
a
hook
before
the
export
happens,
so
cala
having
a
callback
you
can
implement
when
you
create
the
exporter,
so
you
can
add
it
before
the
export,
maybe
it's
cleaner
than
in
everything.
B
C
B
C
C
C
They're.
Borrowing
from
that.
C
I'm
just
I'm
just
thinking
like
so
if
this
becomes
a
spec
thing
like
where,
where
it
is
a
like
a
an
optional
configuration
to
turn
this
on
to
have
spans
time
stamped
at
export
time,
then
it
kind
of
is
unnecessary.
To
do
this.
To
do
this
extension
that
I
need
to
do,
and
so
adding
a
callback
might
not
be
necessary
in
the
long
run
like
for
it
to
to
behave
as
it
is,
but
I'm
just.
C
B
Not
a
video,
I
really.
I
really
prefer
that
you
use
the
default
exporter
or
you
inherit
from
it
just
because
it
improves
the
project
for
anyone.
You
have
your
own
exporter,
you
take
your
class,
but
at
the
end
you
will
end
up
differentiating
from
it.
Yeah.
D
B
You
could
also
have
problems
with
the
with
you
know,
with
a
spec
changing
behind
you,
and
also
the
project
wouldn't
benefit
from
the
pixels
or
bugs
you
could
find.
B
I
think
yeah
I
was
just
thinking
about
the
callback
idea
just
to
not
checking
if
being
open
for
that
class
is
good
or
not.
But
if
you
don't
like
the
callback
idea,
I
yeah
I
will
check
and
we'll
make
it.
I
will
make
it
open.
C
Yeah,
I
think
the
callback
idea
will
just
make
it
a
little
bit
more
messy
in
the
long
run
and-
and
I
was
wondering
like
if
if
one
exporter
should
be
made
extensible
open,
then
maybe
maybe
it's
a
good
idea
to
make
them
all
extensible.
I
don't
know.
B
B
I
will
check
that
that
that
that
what
was
why
I
said
that
I
see
you
guys
yeah,
that
happens,
for
example
with
the
with
the
own
tracer
sdk.
It
must
be
final.
If
you
see
other
languages,
you
see
java,
for
example,
it
says
it's
final
class.
B
And
that's
what
I
think
should
happen
with
also
the
open
telemetry
provided
exporters,
but
I
will
check
that
but
yeah.
I
will
check
that
and
if
you
don't
like
the
callback
thing,
we
can
try
go
with
the
open
one.
I
will
check
with.
A
C
B
Okay,
yeah,
I
yeah.
I
don't
know
about
that.
I
don't
know
other
solutions.
Yeah,
inheriting
or
having
a
callback
would
work.
But
if
you
don't
like
the
callback
yeah,
I
I
will
check
if
it's
possible.
If
not
you
are
not,
you
don't
want
the
callback
right.
B
There
is
also
the
spam
processor
to
add
things
to
the
spams,
but.
A
B
So
it's
just
before
the
export
happens,
so
it's
called.
It
might
be
good
enough.
It
might
be
good
enough
also.
I
am
doing
that,
for
example,
I
have
to
I
have
a
spam
processor.
You
can
have
several
spam
processors,
so
there
is
no
problem
having
another
spam
processor
extra.
Just
to
add
things
to
all
the
expands.
B
My
project
is
a
testing
instrumentation
with
open
telemetry,
so
I'm
instrumenting
tests
in
swift
and
I
am
adding
a
task
for
all
the
expanse
created
when
I
am
running
so
I
am
using
jana
just
a
spam
processor
to
add
that
to
all
the
spams
when
they
start,
but
you
have
a
call
back
also
at
the
end,
so
it
happens
just
previously
to
the
export.
It
happens,
the
end
of
the
spam
processor.
B
So
as
you
as
you
want,
and
also
you
have
also
a
problem
with
that,
is
that,
depending
on
on
what
you
know,
if
you
are
using
the
you
can
also
the
spam,
the
budget
spam
processor,
the
simple,
simple
processor,
they
can
be
calling
the
export
in
different
times.
B
I
don't
know
I
mean
you
can
be
calling
the
export
five
or
ten
seconds
after
the
spam
has
been
really
ended.
I
don't
know
if
that's
what
you
really
want.
C
Yeah
yeah,
so
the
the
the
concern
is
with
the
the
transport
time
between
when
the
span
gets
sent
to
the
yeah.
This
is
yeah.
B
A
B
Yeah,
okay,
then
the
spam
processor
is
not
an
option
for
you,
so
yeah
we'll
we'll
check
about
that
and
and
we'll
if
there
is
no
way
of
making
it
open.
I
will
add
a
I
will
add
a
callback
to
the
constructor,
so
you
can
just
process
it
will
receive
the
an
array
of
expanse.
B
That
is
the
one
that
calls
to
the
exporter,
so
you
can
modify
them
as
you
want.
Okay,
that
that
would
be
the
option
if
the
open
is
not
possible.
Okay
sounds.
C
B
D
Okay
cool,
so
so,
just
to
you
know
for
me
to
understand
bryce
you're,
adding
another
attribute
right.
C
Trying
to
think
if
I
can
find
the,
I
don't
think
I
don't
think
I
know
where
it
is
there.
There
is
a
ticket
discussing
this
as
an
as
a
setting
to
be
added
to
the
spec
for
open
telemetry,
but
I'm
not
sure
where,
where
it's
at
it
was
just
shown
to
me
by
the
team
that
that
asked
me
to
add
this
to
help
them
solve
clock
skills.
C
So
the
problem
is
in
distributed
tracing
that
the
timestamps
that
are
generated
in
traces
on
a
server
are
using
a
different
clock
than
the
one
that
is
on
the
iphone
and
so
and
and
the
problem
grows
even
deeper,
because
the
clock
on
the
iphone
is
controlled
by
the
user,
and
so
it
could
be
set
to
anything
and-
and
that
will
result
in
these
huge
gaps
in
the
the
trace
because
of
these
timestamp
skus,
and
so
using
that
getting
that
export
time
will
help
identify
clock
problems
as
those
traces
get
reported
to
the
get
reported
to
the
collector.
C
I'm
thinking
what
it
does
is
it'll
say
like
here's,
the
time
stamp,
that
was
the
export
timestamp
that
was
set
on
the
device
and
here's
the
receive
timestamp
and
if
they're
too
different,
then
it
can
be
applied
that
offset
difference
can
be
applied
to
all
the
timestamps
in
the
trace
itself.
So
it's
like.
Let's
say
it
was
like
20
minutes,
and
it's
like
you
literally
just
sent
this
and
it
took
20
minutes
to
receive
it.
C
I
didn't
read
the
the
spec
plan
for
it,
but
they
seem
to
think
that
this
will
help
them
identify
clock,
skew.
D
D
So
when
you
do
export
so
you're
adding
the
export
time,
so
you
can,
you
know,
figure
out
the
difference
and
if
you
know
so,
the
create
time
versus
the
star.
The
export
time
on
the
device
will
both
be
skewed
by
the
the
incorrect
setting.
C
Right
yeah,
so
the
problem,
the
problem
is,
you
can't
use
the
create
time
because
there's
no
frame
of
reference
for
the
processing
side
of
things
to
to
figure
out
what
that,
what
that
that
sku
is
using
that
export
timestamp,
the
the
processing
side,
the
side
with
the
the
trusted
time,
the
trusted
clock
can
can
look
at
that
export
timestamp
and
compare
it
to
the
receive
time
because
there's
a
very
small
gap.
That's
that's
probably
you
know
pretty
consistent,
so
it
can.
It
can
adjust.
C
B
Okay
yeah,
so
the
idea
is
having
a
time
that
you
more
or
less
consider
the
same
time
in
server
and
the
device
to
know
the
difference.
There
is
what
gives
you
the
whole
difference
of
all
your
device.
Yeah.
C
C
Yeah,
the
problem
is:
is
that
n2p
isn't
100
reliable?
Just
because
you
know
the
network
conditions
might
not
allow.
You
know
you
try
to
set
it
and
the
network's
down
at
that
moment
or
something
along
those
lines.
C
B
C
I
think
that
I
think
that
the
solution
will
be
kind
of
a
multi-tier
thing
like
yeah,
using
an
ntp
using
this
clock,
skew
correction
stuff,
maybe
there's
some
other
things
that
can
be
done
as
well,
but
I
think
this
is
the
the
lowest
like
the
lowest
hanging
fruit
at
the
moment
is
is
doing
this
sort
of
back
in
calculation.
B
Yeah
yeah
yeah
good
yeah.
Okay,
then
I
will
check
that
and
see
if
that
can
be
done.
B
B
Oh
sorry,
sorry,
0.4,
zeros
or
something
like
that,
so
it
was
the
between
the
first
and
the
second
release.
I
j.
I
changed
the
time
and
just
had
an
error
with
converting
milliseconds
to
make
microseconds
yeah
the
code
said
ms
and
I
thought
it
was
millisecond,
but
it
was
microseconds
yeah.
It
was
really
helpful.
So
I
changed
all
the
ms
thing
and
named
that
microseconds.
B
So
there
won't
be
that
problem
again
yeah.
So
it
was
taking
time
that
they
were
so
big
that
they
were
happening
just
in
the
future,
like
in
1
000,
more.
B
Longer
than
it's
so
yeah
was
really
nice,
but
no
no
one
has
very
long
time
so
yeah,
okay,.
D
Other
exporters
don't
have
this
problem
right,
did
it
none
of
them.
Have
the
micro
like
the
prometheus,
there's
a
few
more
right
in
there
yeah
I
should
I
should
I
didn't
test
that
out.
I
think
the
prometheus
one
also
was
not
when
I
ran
it
a
couple
of
weeks
back
those.
D
D
D
B
B
Okay,
I
will,
I
will
take
a
look
if
I
can
find
was
because
I
also
know
that
that
was
working
when
I
wrote
that
yeah.
B
Definitely,
okay,
yeah!
I
will.
I
will
take
a
look.
If
I
can
find
the
issue,
I
will
try
to
add
it
before
the
release.
D
B
D
Just
yeah.