►
From YouTube: 2022-03-17 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).
B
A
Yeah,
not
bad,
just
it's
working
as
usual.
No
doesn't
seem
like
there's
going
to
be
any
attendees
today.
B
A
Was
how's
your
first
week
as
a
maintainer
life.
B
Yeah,
it's
a
good,
nothing
like
no
major
changes.
It
says
yeah.
A
Yeah,
it's
we
forgot
to
put
the
is
recording
check
because
we
directly
access
the
kind
attribute
and
that's
only
for
sdk
spans,
so.
A
And
it
was
for
these
two
instrumentations,
because
the
previous
release
had
the
interpretations
checking
for
server
spans.
There
was
a
change
for
that
yeah,
so.
A
B
B
B
A
I
see
okay,
cool
cool,
all
good
yeah-
I
guess
something
just
like
not
as
formal
but
like
we
can
just
go
through
so
the
flask
tornado.
A
We
should
probably
do
like
a
hot
fix
release
for
these
two.
It
is
like
block
blocking
people
who
are
trying
to
use.
It
makes
sense.
A
Yeah,
I'm
pretty
sure
we
we
only
need
to
well,
usually
when
we
do
hotfix
releases.
It's
only
for.
B
A
So
I
think
we
bump
everything
by
the
whatever
beta
version,
but.
A
To
bump
open
telemetry
itself
so
like
the
the
core
repo.
A
Cool
is
this
something
that
you're
interested
in
trying
to
do
because
you've
never
released
before.
B
A
Oh
yeah
for
sure
for
sure,
so
usually
we
start
off
with
the
releasing
md
file
in
core.
A
Just
right
here,
yeah
pretty
much
follow
those
instructions.
This.
This
is
only
for
actual
releases.
You
know,
but
you
can
apply
this
for
the
for
the
contrib
repo
as
well
yeah.
So
the
the
good
thing
is
like:
if
you
follow
this
but
you're
confused
then
like
we
actually
have
to
fix
our
readme,
so
it
would
be
it'll
be
good
to
have
any
feedback
that
like
if
you,
if
you
don't
really
understand
anything,
then.
B
A
Yeah
awesome
yeah.
Let
me
know
if
you
have
any
prop
problems,
sure
I
think
in
order
in
in
terms
of
prs
and
everything.
Let
me,
let
me
add
some
of
my
ones.
B
Hey,
I
don't
know.
A
Yeah,
I
didn't
even
know
all
good,
though,
did
I
miss
anything
or
no?
Not
in
particular,
I
don't
know
if
you
both
knew,
but
we
had
a
regression
error
in
our
latest
release.
Two
of.
A
Yeah,
yes,
it
pretty
much
just
relates
to
the
previous
release.
We
had
this
change
in
which
we're
trying
to
access
the
kind
attribute
of
the
span,
but
that
doesn't
exist
on
sdk
on
non-recording
spans
on
the
api,
so
like
yeah
people
are
getting
runtime
errors
for
that.
So
alex
made
the
change
for
whiskey,
but
we
just
forgot
for
the
other
two,
so
we're
trying
to
do
a
hotfix
release.
A
A
Yeah
the
non-recording
span
has
no
attributes.
Sorry,
no,
no,
no
properties
on
it,
so
we're
not
supposed
to
directly
access
fields
on
the
object.
So.
E
Sorry
span
has
there's
like
a
span
interface
in
the
api
right,
yeah.
E
Like
and
get
this
get
spam
contacts
set
attributes.
A
B
A
Yeah,
I
know
so
I
think
I
think
we
were
talking
it's
like
like
if
we
had
like
more
robust,
like
you
know,
api
layer
tests,
you
know
to
catch
this
kind
of
stuff.
I
think
that
would
have
been
definitely
aired
out.
I
think
we
in
terms
of
instrumentations,
like
we've,
been
very
loose
in
like
what's
it
called
like
what
we
can
access
and
everything
it's
just
like
in
the
regular
testing
like
we
just
assumed
that
the
user
has
the
sdk
installed.
You
know
yeah,
so.
E
I
mean
what
is
the
actual
preferred
pattern
here,
because
if
you
do
like
like
this
is
this
is
something
that
type
checking
would
catch
right
like,
but
if
you
do
get
tracer
and
then
tracer
dot,
that
gives
you
an
api
tracer
and
then,
if
you
do,
api,
tracer
dot
start
span,
you'll
get
an
api
span,
which
is
the
one.
That's
missing
all
the
properties,
but
if
you,
if
you
do
want
to
make
an
assumption
like
are
you
just
supposed
to
do?
An
is
instance
check
like
that's?
Not
that's,
not
a
great
api
either
right.
A
E
A
Yeah,
that's
actually
a
good
point.
A
I
think
it's
capture
custom
requests
as
span
attributes.
E
E
A
All
right,
so
this
is
like
a
feature
basically
yeah.
We
can.
B
A
Like
what
kind
of
of
the
the
response
headers
that
we
get
is
captured
as
spam
attributes
and
originally
like
this
is
an
okay
change
by
itself,
but,
like
I
think,
away
commented
that
we
only
wanted
this
for
server
spans.
A
Yeah,
because
of
the
specs,
I
think
it
it
doesn't
imply
that
it's
for
all
types
of
spans.
That's
why
yeah
this
thing
right
here?
A
Okay,
I
guess
from
the
wording
they
took
it
as
if
like
it,
it
is
a.
It
only
applies
for
service
bands.
A
I
don't
explicitly
see
that,
but
like
it
was,
it
was
a
small
enough
change
that
I
think
it
was
like
not
a
huge
deal,
but
I
think
this
access
of
the
kind
was
kind
of
overlooked.
A
E
A
E
Wasn't
there
reasoning
to
make
the
instrumentation
like
the
api
span,
read
only
that
there
was
some
case
where
you
might
not
be
able
to
read
it
or
they
wanted
to
leave
open
in
the
future.
The
possibility
that.
E
E
Hear
me
if
you,
if
you
open
up,
if
you
expand
around
the
kind
check,
spend
all
kind
check
for
server.
B
B
So
why
is
for
propose
that
only
the
whiskey
server
span
remains
as
a
server
span
and
all
other?
You
know
these
frameworks
have
the
servers
like
the
span
kind
of
internal.
So
I
think
that's
the
reason
why
you
know
there's
a
span
can
check
here
because
we
don't
want
to
you
know,
repeat
the
same
attributes
like
same
similar
attributes
in
this.
B
There's
a
possibility
like
if
you
just
use
the
like,
if
you
just
if
you're
just
using
the
flask
instrumentation,
it's
just
our
spam,
but
we
are
using
both
whiskey
and
flask
this.
This
will
be
internal.
I
see.
E
Interesting:
okay,
thanks!
That's
that
that
makes
more
sense.
A
Yeah
sure
so
this
seems
to
be
kind
of
like
an
implementation
detail.
Right
like
it
might
be
like
an
easy
way
to
check
the
double
instrumentation
case,
but
didn't
we
have
some
other
mechanism
to
do
this
as
well.
B
Like
this
is
not
specific
to
like
several
instrumentations,
it's
like
this,
the
same
problem
exists
in
client
instrumentation.
I
I
don't.
I
don't.
You
know
like
remember
how
this
is
being
handled.
You
know
in
other
places,
but
this
is
the
like.
The
reason
the
kind
check
is
here
is
because
you
know
we
don't
want
to
repeat
this.
A
Right
right,
so
I
guess
like,
in
my
opinion
like
to
do
this
properly
like
this
is
kind
of
like
the
hacky
way
right
also.
This
is
like
a
separate
topic
from
like
accessing
attributes,
but
like
for
specifically
this
use
case.
It's
like,
wouldn't
it
be
more
robust
to
like
use
a
mechanism
for
double
instrumentation
than
just
using
this.
B
Yeah,
I
also
had
a
different
question
about
like
this,
this
mechanism
because,
let's
say
like
whiskey
instrument
like
their
two
instrumentation,
whiskey
and
flask,
and
then
some
some
some
middleware
like
some
something
in
the
middle
has
added.
You
know
some
headers
right
and
then
you
want.
You
expect
them
to
be
captured
in
the
you
know
the
flask
instrumentation,
but
this
check
skips
that,
like
it
does
not
add
them
here,
right
yeah.
I
had
like
that
doubt,
but
there
was
no.
You
know
clear
answer
for
that.
D
B
Like
we
can,
we
can
try
to
do
this
in
more
hours.
Instead
of
this.
E
Yeah,
that
makes
sense,
but
one
other
thing
I
was
going
to
say:
is
it
also
in
the
api
span?
There's
a
is
recording
check
and
the
actually
the
other
spot
in
the
same
vr,
where
they're
checking
the
span
kind
checks.
If
it's
a
recording
spin
beforehand.
A
Yeah,
so
the
we
we
did
it
for
whiskey.
There's
two
pr's
one,
sorry,
three
pr!
It's
one
for
flask,
one
for
tornado,
one
for
whiskey
and
we
did
the
is
recording
for
check
for
whiskey.
But
we
forgot
for
flask
and
tornado.
That's
why
those
are
the
ones
that
need
a
hot
fix.
A
B
E
In
terms
of
like
preventing
this
kind
of
thing
in
the
future,
because
I
think
I've
seen
this
a
few
different
ways
where
we
just
assume
we
have
an
sdk
object,
and
it's
not
necessarily
true,
I
mean
if
we,
if
we
were
able
to
run
my
pi
or
something
like
that
on
this
code,
it
should
be
able
to
to
catch
stuff
like
that.
E
It's
obviously
going
to
make
our
code
a
little
bit
more
verbose,
because
we'll
have
to
do
right,
like
typing,
is
instance,
checked
or
or
ignored
my
pie,
or
something
like
that
which,
which
is
okay,
I
think
where
it
needed,
but
I
don't
know
if
we
want
to
feel
like
that's
something.
We've
been
pushing
off
for
a
long
time,
yeah,
but
catch
stuff
like
this
right.
B
Yeah,
I
agree
with
that,
and
other
thing
is
that,
like
you
know
you,
so
all
these
instrumentations,
you
know
rely
on
this
test
base
that
we
have,
which
is
actually
setting
up
the
sdk.
So
like
one
quick
way
to
you
know
see
how
like,
if
the.
If,
if
there
are
any
errors
like
attribute
errors,
we
can
set
an
op
pressure
provider
and
see
you
know,
but
that's
like
a
quick
option
to
see.
If
there
are
any
errors
but
like
we
will
have
to
update
tests
to
make
sure
that
they
don't
fail.
E
Yeah
yeah,
I
think
that's
a
good
idea,
one
other
one
thing
I
was
going
to
say
is:
I
think,
if
diego
was
here,
he
would
mention
the
like
error
handling
stuff.
I
think
in
this
case
this
is
like
something
we
wouldn't
want
to
catch
with
that
error
handling.
But
what
do
you
guys
think
you
agree.
E
A
E
Like
in
like
the
most
magical,
like
far
extending
version
of
it,
it
would
basically
make
make
this
thing
all
these
sdk
things
behave
in
a
way
that
it's
impossible
for
the
instrumentation
to
raise
an
error,
but
I
think
I
think
this
behavior
is
something
that
we
shouldn't
even
bother
like
it's
a
programming
error
in
the
instrumentation.
I
don't
think.
B
Agree
with
that,
like
it's,
it's
our
mistake
right,
it's
not
something!
You
know
user
like
incorrectly
configuring
or
you
know
incorrectly
using
it.
It's
our
mistake,
programming,
programmatically,
so
yeah.
A
But
like
the
whatever
magic
or
implementation
way
that
diego
was
thinking
of
doing
this
error
handling,
would
it
have
buried
this
error.
A
E
Yeah
like
like,
if
you
look
at
their
end
up
handling
principles,
it's
like
oh
instrumentation,
shouldn't
crash
the
app,
but
it's
just
difficult
in
python,
where
you
know
people
aren't
doing
a
lot
of
static
checking
so.
A
E
E
B
B
We
are
seeing
this
error
like
they
are.
They
don't
have
the
sdk.
C
E
A
If
this
is
this
like
this,
we
didn't
do
anything
to
make
this
happen
right.
It's
like
it's
just
an
like,
like
python,
just
through
an
error
right.
E
A
E
E
And
it
shouldn't
cause
like
503s
or
something
like
that
or
or
whatever
right.
So
I
think
logging
is
totally
fine
and
if
we're
using
a
logger
we
can
they
can
silence
the
log
messages,
meaning
like
if
we
cut
a
bad
release.
It's
not
gonna
be
like
a
cause,
a
huge
outage
for
all
these
people
if
they
upgrade
their
thing
right,
yeah
you're,
just.
A
A
Yeah
cool
right
so.
A
That
is
a
good
point
too.
So
you
see
all
of
these
like
issues
right
here
that
alex
created.
A
That
could
also
be
a
good
sanity
check,
as
well
as
part
of
this,
making
sure
that,
like
if
some
error
happens,
like
those
things
don't
have
the
we
don't
crash
the
app
as
well,
so
we
actually
did
something
similar
like
alex,
and
I,
when
we
were
like
releasing
our
stable
version,
we
literally
had
to
go
through
the
entire
like
api,
surface
and
audit
each
thing.
So.
Similarly,
I
think
we
can
do
that
for
instrumentations,
eventually.
A
Yeah,
oh,
we
already
have
like
audit
okay
yeah.
So
that's
pretty
much.
I
think
sukanth.
I
don't
know
if
you
made
that
issue
already.
Oh
sorry,
that's
a
different
issue
but
yeah
what
I
wrote
down
about
the
you
know
modified
test-based
details.
Possibly
that
could
be
part
of
this
work
right.
A
A
Yeah,
okay,
cool.
Does
that
make
sense
to
you
guys,
like
anything
else,
are
we
gonna
talk
about
for
that.
A
E
A
Awesome
cool
cool
for
prs.
A
I
put
up
one
recently
about
capturing
exception,
information
for
vlog
attributes.
Yes,
sir,
thanks
for
taking
a
look
at
this,
does
it
seem
like
we
do?
Have
an
open
ended
question.
A
A
Information
is
the
the
exception
of
stack
trace
as
one
of
the
the
attributes
of
the
logs
and
the
specs
doesn't
really
specify
like
how
the
format
of
the
string
that
is
being
captured,
so
kind
of
suggest
like
using
the
the
traceback
module,
but
it
prints
it
out
in
like
a
really
ugly
kind
of
thing
like
this,
you
see
like
there's
a
bunch
of
like
white
spaces
in
between
a
lot
of
like.
E
A
A
Right,
yeah
yeah,
so,
like
that's
fine
like
it
contains
all
this,
the
information
in
a
normal
stack
trace.
I'm
just
saying
like
like
in
my
implementation
right
now,
I'm
like
printing
it
nicely.
A
You
know
like
I'm
like
making
a
dictionary
and
like
printing
it
out
as
json,
so
like
users
can
like,
if
we
specify
this
format
like
you,
just
can
easily
parse
it.
You
know
and
find
exactly
what
they're
looking
for.
So
it's
just
like
a.
B
A
Print
pretty
much
versus
like
this,
this
weirdness
right
here
so
like
we
were
just
talking
about
it
and
it's
like
so
most
other
sdks
use
the
like,
whatever
stack,
trace
or
traceback
library
right
and
they
just
print
out.
A
Just
totally
fine,
but
I
think
did
bring
up
a
good
point
that
you
can
just
read
it
right
here.
E
I
think
like
so
the
use
case
for
this
is
like
I
load
up
my
logs
and
I
see
the
trace
back
right.
Yeah
like
there's
a
there's,
a
problem,
there's
like
an
exception
right,
so
it
I
mean
if
they
want
it
formatted
in
a
certain
way
like
it
doesn't
seem
super
useful
unless
we
formalize
that
anyway.
So
this
makes
it
sort
of
pretty
printed
as
a
string
verse
like
json,
encoded.
E
Yeah
I
mean,
I
think
for
sorry,
I
I'm
not.
This
is
like
not
something
I've
thought
about
a
lot,
but
I
think
just
putting
it
as
a
string
for
now
probably
makes
the
most
sense
just
because
like
if
we,
if
we
did
do
it
as
json,
we
could
change
it
and
then
that
would
break
people,
whereas
if
we're
just
doing
a
string
or
whatever
people
aren't
going
to
be
likely
to
do
that.
A
That
is
true,
I
think
like,
if,
if
we
in
the
future,
if
we
ever
specify
a
format
like
we'll
break,
people
anyways
but
like
it's
like
it's
like
going
from
json
to
string
is
is
uglier.
I
guess
you
know
like.
Oh,
we
ruined
their
prettiness.
B
A
If,
if
they
just
already
started
with
string,
it's
like
oh
okay,
anything
is
better
than
this,
this
right,
so
okay.
So
what
does
this
backstay
on
this
they're?
Not
very,
I
don't
think
they
specified
anything
about
the
format
they
just
said
that
like,
if
you
specify
like
some
way
to
do
it
just
be
sure
to
inform
users
of
what
it
looks
like.
A
D
B
E
A
E
Yeah,
I
guess
that's
fine.
I
think
that
I
think
that
probably
makes
sense
to
start.
I
imagine
there's
some
like
stack
trace
format
available
somewhere.
I
think,
like
I
know
there
is
this.
E
E
E
More
appropriate
for
aristide
choices,
but
right
right.
A
Okay,
cool
srikant
is
that
does
that
make
sense
well
like
what
we're
going
to
be
doing.
B
B
Yeah
I
was
looking
at
the
pr
and
I
thought,
maybe
you
know
bring
up
this.
So
if
you
look
at
the
the
link,
we
are
right
935,
so
the
other
you
know
the
change
is
that
he's
pumping
the
like?
The
author
is
pumping
the
the
minimum
support
version
to
three.
So
what
so
far
we
are
only
supporting
like
one
point
x.
B
Other
is,
you
know,
bumping
it,
but
you
know
like
if
you
look
at
the
requirements,
he's
pumping
it
for
the
three
point
x
so
like
what
does
it
mean
for
the
you
know,
people
who
are
already
using
this
instrumentation
like
how?
How
do
we
handle
it,
for
you
know,
other
versions
that
are
there.
This
is
as
an
example,
but
in
general,
this
issue
will
be
in
this.
This
will
be
an
issue
in
other
instrumentations.
Also
so
like
I
I
wanted
to.
B
You
know,
bring
it
up
and
here
your
ideas
about
how
do
we
want
to
handle
multiple,
like
instrumentation
library,
like
multiple
major
versions
of
the
library.
D
A
I
think
okay,
so
just
to
clarify
your
questions
like
this
guy
is
bumping
up
the
supported
instrumented
version
from
one
point
x
to
3.6
to
three
point
x
right.
B
B
B
In
the
same
instrumentation,
if
it
wasn't
going
to,
you
know
make
that
possible.
We
intend
to
release
like
the
different
instrumentation
libraries,
but
for
the
sake
of
convenience,
make
it
easier
for
the
users
we
went
ahead
with.
You
know
trying
to
accommodate
the
changes
in
same
instrumentation
library,
but
it's
not
always
the
case
that
it
you
know
we
can
do
that.
B
In
this
case
it
looks
like
you
know,
you
can't
have
the
same
thing
like
multiple
versions
here,
yeah.
A
Right,
I
think
we
did
bring
that
up
too.
I
don't
know
correct
me
if
I'm
wrong,
but
I
remember
discussing
that,
like
our
message
to
customers
was
just
to
like
we
support
something
up
to
a
certain
version
and
you
can
keep
using
that
instrumentation
library
to
that
version,
but
like
if
you
want,
like
the
newer
features
of
the
instrumented
library,
you
just
have
to
upgrade.
A
C
B
I
don't,
I
don't
remember
that
discussion.
A
Okay,
well,
maybe
I'm
just
making
stuff
up
all
right.
A
Sorry
yeah
the
horror
stories,
yeah
yeah,
maybe
maybe
this
is
like.
Maybe
I
talked
about
this
within
microsoft.
So,
like
that's
like
nobody
understands
it's
like
yeah,
I
think
like
like,
like
the
stance
that
we
were
taking
is
like
you
just
you
just
upgrade.
If
you
want
like
the
newest
stuff,
I
believe.
A
B
Yeah
I
mean
like
with
people
people
like
who
started
relying
on
the
instrumentation.
Let's
say
in
in
one
point:
let's
say
now:
we
bump
up
right
like
we
are,
you
know
we
tell
them
to.
B
You
know,
upgrade
the
libraries,
but
they
can't
they
they.
They
might
have
some
reasons
not
to
do
that,
and
maybe
there's
some
bug
that
they
want
to
get
it
fixed
in
the
old
versions,
but
now
they
cannot
do
anything
either
like
if
they
can
fix
it
and
they
can't
pump
it
like
that.
So
for
them
like
that
situation
is,
was
so
like
trying
to
think
about
the
you
know
these
scenarios
like
how
does
it
affect
the
end
user,
yeah.
E
A
A
Okay
right
so
yeah,
maybe
we
need
to
discuss
this.
Maybe
other
people
have
ideas
as
well
so
circa?
Maybe
you
could
bring
this
up
either
in
the
slack
channel
or
we
could
talk
about
this
in
the
next
week's
meeting.
B
Yeah,
just
wanted
to
you
know,
start
some
something
so
that
you
know
this.
Would
this
I
I
can
see
how
this
can
become
a
problem
in
future
also.
I
just
wanted
to
start
with
now,
so
that
we
will
have
some.
A
Consensus
awesome
yeah
thanks
all
right,
any
other
issues
guys
to
bring
up.
E
A
Yeah,
well,
I
don't
think
we
have
any
more
agenda
items,
so
we
can
just
do
that
so
condolences.
If
there's
anything
else,
you
want
to
talk
about.
E
So
we
did
have
one
come
in,
which
was,
I
haven't,
really
looked
at
it
yet,
but
I
think
it's
basically
remember.
I
had
a
pr
to
reset
the
points.
Yes,
let's
get
something
similar
for
histogram,
where
the
sum
wasn't
being
reset
and
they
also
included
vr.
So
I'll,
take
a
look
at
that,
but.
A
A
E
Yeah,
I
think
I
have
a
feeling
this
person
I've
seen
their
name
before
I
think
they're
in
active
in
like
another
sig.
Maybe
I
see
yeah.
A
B
E
They,
I
think
they
work
with
dyna
trace
yeah,
which
is
one
of
the
other
delta
cool,
so
yeah
I'll
take
a
look
at
the
pr
but
awesome
other
than
that.
I've
been
looking
at
things
that
there's
been
a
lot
of
changes
to
the
spec
lately
and
there's
some
things
we're
missing.
If
you
go
back
to
the
project
board,.
E
E
A
E
E
E
Yeah
yeah,
so
that's
that
one
and
then
default
is
you
know,
basically,
it's
just
a
explicit
way
to
say
use
the
default
aggregation
for
for
the
instrument.
B
A
E
Other
than
that,
I
have
a
few
like
discussion,
things
that
I
opened
previously.
There
is
one
that
was
for
the
multi-instrument
observable
callback
stuff,
which
was
to
get
merged.
So
sorry,
it's
not
actually
in
discussion.
It's
in
the
project
board
they're
just
issues.
Oh.
E
Yeah
and
diego
was
some
comments-
I
don't
know
if
either
of
you
guys
have
any
thoughts
on
this,
but
it
would
be
good
to
make
a
decision
and
and
start
implementing
it.
A
Yeah
I'll
have
to
read
this
again.
It's
been,
it's
been
a
minute.
E
Just
add
that
yeah
and
I'm
probably
gonna,
make
a
few
more
issues
today
for
other
things
that
that
I
notice
are
missing
or
we
should
change.
Yes,.
A
Hey
aaron,
how
are
you,
how
are
you
finding
these?
Are
you
just
going
through
a
code
base
or
like
do
you
have.
E
E
Oh
so
so
some
some
of
these
are
just
like
spec
things
that
we
don't
have
you
talk
about,
like
the
bugs
I
was
fixing
last
week,
yeah,
oh
yeah,
so
I
was,
I
was
actually
playing
around
with
stuff,
so
the
cool,
the
metrics
sdk
works
and
you're
like.
E
Doesn't
work
properly,
yeah
yeah,
I
just
I
was
just
trying
stuff
out,
so
it's
actually
pretty
cool.
A
lot
of
things
are
working
now
I
think
we
have
we
could
we
could
be
making
some
sort
of
experimental
announcement
asking
people
to
try
it
out
or
update
our
readmes
and
stuff.
At
this
point,
I
think
so.
A
Yeah
I've
been
already
trying
to
get
people
on
this
because,
like
like
you,
similarly
like,
I
think
people
would
be
able
to
find
out
more
issues
by
using
it
yeah.
A
Than
just
visibly
going
through
the
spec
and
finding
what
we're
missing
so
yeah
cool
sounds
good.
E
A
Awesome,
did
you
want
to
make
issues
for
those.
C
A
E
E
A
B
A
The
full
hour
there's
only
three
of
us:
we
love
talking
yeah,
actually,
okay,
cool.
If
not
because
I
get
seven
minutes
back
and
I'll
see
you
guys
next
week.