►
From YouTube: 7 Oct 2022 Node.js Node-API Team 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
All
right,
hello,
everybody
and
welcome
to
this
week's
node
API
meeting
before
we
get
started
with
our
regular
agenda.
Does
anyone
have
any
announcements
they
would
like
to
make
nope
all
right?
Then
let
me
share
my
screen
and
we'll
get
started
with
our
normal
agenda.
A
I
did
take
a
look
at
this
one
here
and
I
just
wanted
to
sort
of
clarify
my
thinking
in
that.
What
we
are
testing
is
that
the
actual
cancel
call
fails
is.
A
B
A
Okay,
so
what
we
have
here
is
you're
using
a
conditional
variable
so
that
when
you
cue
the
work
to
start-
and
it
begins
executing
here-
so
you
cue
it
to
start
and
then
you
wait
for
this
ready
to
become
true,
which
initially
is
false
and
it
gets
set
to
True
here
and
then
you
notify
so
that
this
continues
and
then
you
cancel
and
then
this
ends
up
throwing
yep,
okay,
yep.
That
looks
good
to
me.
A
So
then,
in
this
case,
when
the
cancel
fails
is
the
normal
on
okay,
then
supposed
to
execute
yeah,
so
I
think
normally
we
I
think
in
the
previous
test.
I
threw
in
their
cases,
but
then
I
think
that
conflates
the
sort
of
the
JavaScript
exceptions
we're
going
to
catch
from
JSI,
so
I
just
kind
of
let
them
like
gotcha,
okay,
yeah
well,
and
then
you
have
this
try
catch
here.
A
So
you
try
to
cancel
it
and
you
would
get
this
throw
here
and
then,
if
it
catches
task
failed
to
true,
and
then
you
expect
it
to
be
true.
Yes,
this
looks
good
to
me.
A
A
Overloads
yeah,
so
I
didn't
get
to
get
to
this
this
week,
but
I'll
I'll
make
time
for
this
to
get
to
go
over
this
again.
Thanks
no
problem
at
all
and
I
think
that
is,
it
is
there
anything
else
that
you
need
help
with
or
anything
I
think
that
should
be
good
for
now.
Yeah
thanks
all
right,
then,
and
moving
on
we
have
this
better
coverage
of
node,
API
and
YouTube.
A
A
Does
anyone
have
else?
Does
anyone
else
have
anything
that
they
would
like
to
add
or
discuss
regarding
this
issue?
A
No,
all
right,
then
we
will
move
on.
We
have
ADD
list
of
non-c
language
bindings.
A
I
can't
see
the
participants,
so
if
he
ends
up
joining
later,
can
someone,
let
me
know,
but
yeah
Nicola
is
working
on
this
and
we
had
discussed
so
create
a
new
page
in
the
repo
like
a
new
markdown
page
and
then
a
link
to
that
page
in
the
a
core
API
docs,
so
that
it'll
be
easier
for
us
to
maintain
and
edit
in
the
future.
B
Stay
on
the
same
note,
I
was
also
wondered
lately.
Then
somebody
asked
me
about
an
old
API.
Is
it
kind
of
somewhere
at
least
a
police,
no
DPA
bindings
to
different
JavaScript
engines?
I
know
that,
like
we
had
something
for
JavaScript,
we
have
I,
think
I
think
we
discussed
this
multiple
times
but
I
wonder
if
we
have
this
list
somewhere
idea.
A
Yeah
we
we
discussed
the
other
engines
when
we
like,
add
or
modify
the
API
just
to
make
sure
that
the
engines
would
support
it,
but
I'm
not
exactly
sure
we
have
a
list
of
those
bindings,
so
I
think
it's
worthwhile
just
to
mention
that
that
it
should
be
if
we
could
add
the
list
of
those
other
JavaScript
engine
bindings
to
this
documentation
as
well.
B
All
right
I
would
recommend
to
have
different
different
documentation.
Pages
it'll
be
like
nice,
oh
I,
I,
don't
know,
maybe
maybe
it's
the
same.
I
don't
know,
but
just
kind
of
yes
from
one.
It's
if
you
like,
it's
like
node
API
is
like
some
kind
of
layer
right
like
it's
API,
but
on
one
side
we
have
like
different
orientations
of
pcpi
another
side.
We
have
different
users
of
this
API.
If
you
like,
like.
A
I
think
that's
fair,
all
right.
Okay,
anything
else
regarding
this
issue.
A
A
Let
me
always
go
here
and
sort
of
look
what
we
got
going
on
almost
nine
nine,
almost
nine
million
weekly
and
then
we'll
eventually
reach
the
10
million
weekly
and
then
Milestone
like
I
mentioned
last
week.
It
may
be
beneficial
to
work
on
a
new
release.
A
There
are
some
PRS
that
have
been
merged
or
are
waiting
to
be
merged.
That
fix
some
issues,
so
it
may
be
beneficial
to
to
start
on
that.
Let's
create
an
issue
for
that.
A
A
All
right
and
I
guess
that
would
end
up
being
one
zero
Maybe.
B
A
A
A
All
right-
let's
just
submit
this,
so
that
we
can
keep
a
track
of
it
all
right
and
welcome
back
here.
Yep
almost
set
the
9
million,
so
things
are
going
well
enable
debug
testing
for
add-ons.
This
is
work
that
Michael's
been
working
on,
but
I,
don't
think,
there's
been
any
additional
work
to
this.
Let's
just
open
the
issue
up.
A
So
then
moving
on
stale
issues
to
discuss
all
right,
this
one,
this
list
has
been
growing
longer
and
longer,
so
we
have
Unbound
memory
usage.
Yes,
this
was
the
issue
where
creating
a
bunch
of
object
wraps
in
a
loop
without
doing
a
the
next
UV
cycle
or
whatever
event
Loop
cycle,
those
objects
never
get
freed
and
we
discuss
this.
A
And
then
there
is,
this
is
a
PR
that
legendcast
has
been
working
on
as
well
as
I
think
you
Vladimir,
or
at
least
the
two
PRS
are
interlinked
because
they
both
work.
They
both
touch
the
finalizers
and
I.
Don't
believe
legenda
cast
is
on
this
call,
so
what
we
can
do
is
just.
A
Nope
nothing
new.
This
is
something
that
Michael's
posted
almost
two
weeks
ago,
but
I
guess
next
time
when
the
gender
cast
is
on
the
call
we'll
be
able
to
discuss
that
with
him.
Yeah.
B
He
said
a
few
weeks
ago
that
it
seems
like
there
is
some
another
issue
and
he
wants
to
create
another
PR
so
and
complete
another
PR.
Before
he
gets
back
to
this
one
I
see
okay,
he
was
talking
about
adding
some
additional
unit
tests.
He
was
thinking
that
there
are
some
test
cases
missing
right.
A
Okay,
so,
hopefully
he'd
be
able
to
clarify
that
in
next
week's
call,
so
that
we
can
see
what
the
status
is
on
on
this
fix.
B
Yeah,
this
work
kind
of
depends
on
what
Chen
Chong
is
supposed
to
do
and
and
then
it
depends
on
features
because
what
we
decided
last
time,
instead
of
adding
a
complete
new
apis,
like
my
PR,
was
doing,
we
simply
said
we
will
keep
exactly
existing
apis.
But
we'll
have
a
feature
and
feature
will
just
simply
say:
all
finalizers
must
be
run
from
from
garbage
clothes.
So
Tingley
was
saying
people
I
mean
developers
not
supposed
to
call
JavaScript
from
their
finalizers.
A
Right
so
the
pr
that
adds
the
feature
Flags.
Is
this
42
557
we're
dealing
with
allowing
references
to
any
value
types,
so
we're
thinking
that,
once
the
feature
ability
has
been
implemented,
then
you'd
be
able
to
add
a
feature
for
when
the
finalizer
should
run.
Is
that
my
understanding?
Yes,.
B
So
because
my
PR
was
originally
about
adding
a
bunch
of
new
apis,
as
you
see
like
six
of
them
so
now
we're
discussing
like,
let's
not
add
them,
but
rather
have
a
feature
which
controls
this
Behavior.
The
module.
A
Right:
okay:
we
can
put
that
as
a
comment,
so.
A
B
We
currently
think
about
this
feature
flag
that
we'll
have
a
predefined
feature
set
for
each
node,
API
version
plus
we'll
have
ability
per
module
people
to
override
this
set
of
feature
Flags.
B
Okay,
that's
the
way
I.
Think
of
this
override
is
going
to
happen
in
the
Jeep
file.
People
need
would
need
to
provide
a
special
definition
saying
that
an
API
has
custom
features,
and
it
means
they
would
need
to
provide
a
function
with
a
certain
name,
a
certain
signature,
and
it
should
work
this
way
so,
depending
on
this
kind
of
macro
definition,
we'll
choose
either
to
use
this
user
function
to
get
the
set
of
features
or
use
default
function
to
get
set
of
features
in
the
inside
of
their
custom
functions.
B
A
All
right
so
all
right,
API,
Behavior
or.
B
A
To
add
so,
what's
up
okay,
so
we
have
that
all
right,
so
that's
it
for
the
object,
wrap
native
objects,
then
we
have
signal,
does
not
trigger
the
thread.
Safe
function,
I've
actually
created
a
PR
for
this
one.
A
A
I'm
worried,
though,
that
we
were
talking
about
thread
safety
earlier
today,
I'm
wondering
if
one
of
these
tests
are
hanging
no
still
compiling
okay.
Well,
we
can
just
quickly
go
over
the
implementation,
real
quick,
so
what
we
had
discussed
for
what
the
signals
should
do
is
trigger
the
progress
callback
with
with
no
data,
so
what
I
saw
in
one
of
the
existing
tests?
A
B
A
Should
do
so
I
looked
at
the
send
implementation,
which
is,
if
we
go
down
I
think
it
is
yes,
so
here's
the
send
implementation,
which
just
does
send
progress,
data
and
count.
So
what
I
changed
was
for
Signal.
It
also
does
the
send
progress
for
null
pointer
and
zero,
and
this
is
fixed
for
both
the
async
progress,
Q
worker,
as
well
as
the
async
progress
worker
and
I
modified
the
the
tests.
A
So
this
is
the
async
progress,
Q
worker,
where
before
actually
doing
all
of
the
the
sending
of
the
progress
there's
an
initial
signal
that
gets
sent
and
what
we
check
is
and
I
also
added.
This
test
case
count,
and
this
gets
incremented
on
every
call
to
on
progress.
So
on
the
first
call
to
on
progress.
If
the
count
is
not
zero,
which
it
should
be
because
signal
has
zero.
B
A
It
would
set
an
error
and
then
likewise
in
the
async
progress
worker,
I
left
the
block
of
code.
That
was
allowing
you
to
do,
send
null
pointer
and
then
I
just
added
a
new
call
for
signal
and
modify
the
test
here
so
that
if
the
on
the
first
two
calls,
if
the
count
is
not
zero,
then
you
have
an
error.
Otherwise,
on
all
of
the
subsequent
calls,
the
count
should
be
one,
and
that
is
basically
what
this
PR
is
doing,
and
there
was
a
hang
on
one
of
the
pr
runs.
B
Yeah
so
basically
like
a
general
notes
here
about
to
then
they
use
this
condition
variable.
You
can't
just
simply
say
wait
because,
according
to
a
special
opposing
standard,
it
can
have
sporadic
wake
UPS.
Yes,.
A
A
Yes,
that
is
what
the
previous
there
was.
Another
thread
say
function
test
case
that
was
hanging,
and
it
was
because
it
didn't
have
that
method
to
to
call
so
I
think
it
needs
to
be
worked
similarly
to
how
Jack
did
it
in
his
failed
task,
cancellations
where
he
has
this
right.
This
here,
yeah.
B
A
That
makes
sense
to
me
so
I'll
modify
this
PR
to
use
that
method.
Another.
B
Kind
of
quick
note
is
about,
like
you
have
I've
seen
the
YouTube
new
function,
name,
which
has
underscore
in
the
end
of
name
all
right,
I
wonder
if
we
can,
we
can
don't
do
it
because
it's
I
think
it's
not
not
exactly
kind
of
coding.
A
A
It's
just
that
I
did
not
add
that
method.
It
was
already
there.
B
A
I
kind
of
just
the
only
major
change
that
I
did
was
you
know
these
four
lines
here?
These
four
lines
are
the
tests,
so
I
would
be
a
little
I,
don't
necessarily
want
to
change
it,
because
it's
kind
of
already
there
but
I
think
we.
A
Right
right,
yeah,
okay,
it
was
good
so
I
needed
to
comment
on
this
block
here.
A
B
Whatever
we
do
in
this
function,
they
will
start
to
Loop
it
for
you.
So
if
it
has
any
sporadic
wake
UPS,
they
will
do
it.
Do
it
correctly.
A
Right,
okay,
so
that
is
that,
yes,
we
were
on
this
stale
issue,
so
this
PR
should
address
that
still
issue
and,
as
we
had
another
issue
as
well,
this
const
error.
A
This
person
was
trying
to
use
the
signal
and
they
were
getting
a
compilation
error,
but
that
is
also
going
that's
now
fixed
with
with
that
change.
So
this
will
also
be
resolved.
A
Okay,
I
think,
that's
it
for
that
stale
issue,
then
moving
on.
A
We
have
should
handle
exceptions.
This
is
one
that
we
had
discussed
in
a
previous
meeting.
A
A
Yeah
I
think
and
I
guess
maybe
next
week
we
can
discuss
a
little
more
about
it,
then,
because
I
just
would
be
a
little
cognizant
of
introducing
a
lot
of
changes.
If
that
is
the
the
the
path
that
we
want
to
go
is
changing
the
the
default
behavior,
but
I
guess
we'll
just
discuss
that
next
week.
A
A
Let
me
make
sure
that
there's
no
comments
or
anything
nope,
okay,
so
that's
just
waiting
to
be
reviewed.
Then
we
have
this
one.
Okay,
I
tried
to
take
a
look
into
this
one,
just
to
understand
again
what
the
person
is
complaining
about,
and
now
I've
I
have
an
understanding
of
the
complaint,
but
I
really
don't
understand.
I,
don't
know
how
to
how
to
resolve
it
or
anything.
So
what
they
have
is
an
object,
wrap
that
holds
a
thread
say
function
and
in
the
destructor
of
the
object.
A
They
want
to
release
the
threats
they
function
and
they
are
saying
that,
under
certain
conditions,
it's
possible
that
node
would
destroy.
The
thread
say
function
before
it
destroys
the
objects
of
the
add-on.
So,
since
the
thread
say
functions
are
already
destroyed,
calling
this
ends
up
making
a
crash.
B
A
Issue
but
I'm
really
not
sure
how
to
go
about
a
solution
for
this.
A
A
That's
really
all
that
I
have
for
this
issue.
Does
anyone
have
any
thoughts
about
this?
Otherwise
we
can.
B
Just
how
deep,
how
typically
sorry
I
don't
know
much
about
the
thread
say
functions
here?
How
is
it
typically
being
destroyed?
So
it's
typical.
A
Scenario
in
in
my
like
in
my
understandings,
you
are
supposed
to
release
the
thread
say
function
when
the
thread
that
is
using
it
is
finished
so
they're
trying
to
link
the
the
concept
of
the
thread
is
finished,
using
it
with
the
destruction
of
the
object
and
I'm,
not
sure,
if
that's
the
best
way
to
to
handle
that,
because
they're,
because
that
that's
not
the
same
time
span,
you
know
like
a
thread
being
finished
with
it
and
the
object
being
destroyed.
A
B
A
A
B
I
would
say
in
general,
in
modern,
C
plus,
plus
cooling
release
function
explicitly
is
a
bad
practice.
It's
because
people
may
have
used
exceptions
and
the
release
function
may
never
be
called
if
exception,
arrays.
So
technically,
typical
approach,
you're
using
some
value
classes,
air
II
restore
acquisition
is
initialization
right.
So
thank
you.
The
name
of
some
object
goes
out
of
the
scoreboard.
Would
call
this
at
least
for
you.
B
I,
don't
have
any
kind
of
concrete
recommendations
here.
It
just
feels
wrong
that
we
have
to
call
at
least
explicit.
A
Yeah,
the
the
reason
you
have
to
call
release
on
a
thread.
Save
function
is
because,
when
you
create
the
threat
say
function,
you
can
say
how
many
threads
are
using
it,
and
you
can
also
increase
that
number.
Using
this
acquire
method.
B
I
guess,
like
you
know,
C
plus
plus
wrapper,
so
we
could
potentially
introduce
some
kind
of
special
thread,
say
function
holder
which
will
be
kind
of
a
special
class
which
in
Constructor
will
say
acquire
and
then
instructor
will
say
the
release.
So
in
this
case
it
will
be
no
kind
of
roll
calls
to
release
at
all.
A
I
see
I
think
we
do
not
do
that,
because.
A
I
mean
at
least
right
now
the
thread
say
function
class.
All
it
does
is
just
hold
the
value
of
the
napi
thread.
Safe
function,
value
like
it's,
basically,
just
a
wrapper
on
on
holding
that
value.
This.
A
A
I
know
it
came
up
in
discussion
when
the
API
was
originally
created.
I
think
it
just
happened
to
deal
with.
You
know
if
we
did
put
a
choir
and
release
inside
the
the
copy
and
Destructor
that
there
may
be
some
use
cases
where
those
objects
are
being
copied
where
it
doesn't
necessarily
need
to
acquire
a
new
thread
say
function
or
something
like
that.
I
think
it
was
so
that
the
our
implementation
would
be
simple
so
that
we
don't
make
any
assumptions
on
how
the
user
is
trying
to
use
the
thread
safe
function.
B
B
Of
pattern,
and
pretty
much
all
whatever
like
thread
say,
function
provides
a
little
bit
just
kind
of
smart
pointer
if
you
like,
for
on
top
of
it
so
with
different
class,
so
people
can,
if
they
ever
want
to
do
some
interesting
scenarios
like
this.
Maybe
we
can,
it
will
be,
could
be
one
of
our
recommendations
to
the
corner,
at
least
directly,
like
you
just
special
smart
pointer,
holder.
A
Value
right,
I
see
I'm
wondering,
though,
if
it
would
still
end
up
having
if
it
would
still
like
the
the
like.
Have
the
problem,
though
so
imagine
this
now
is
not
a
threat
say
function,
but
it
is
that
that
holder
that
you're
talking
about
so
then
all
that's
really
different
is
that
this
release
would
be
done
inside
this
Destructor
itself,
so
that
the
user
wouldn't
have
to
do
it.
But
I
still
think
that
would
end
up
causing
this
issue
where
the
the
object
is
being
destroyed.
A
B
Definitely
we,
we
must
be
some
good
good
Solutions
there
and
multi-threading
is
generally
difficult,
like
even
in
C
plus
plus
standard,
say
now.
I
recommend
no,
don't
use
subscribe,
use,
J
thread,
because
there
are
some
issues
with
joining
threads
and
they
reduce
new
new
type.
B
In
both
cases
we
see
free
environment
first,
so
it's
possible
that
we
have
some
like,
like
you
said,
maybe
the
functions
at
least
before
of
this
object.
B
A
A
B
You
know
what
I
mean
like
if
we
should
only
destroy
it.
If
ref
count
goes
to
zero
and
rather
than.
A
B
You're
saying
and
maybe
do
talk
in
two
passes
like
imagine
like
first
pass,
we
just
we're
trying
to
do
our
best,
like
always
have
counted
stuff
right
and
if
this
first
possible,
with
different
regular
destruction,
does
not
produce
still.
We
still
have
something
left
over
then.
Only
after
that
we
do
this
deletion.
So
imagine
if
we
freed
up
all
these
objects
in
two
passes.
Right,
like
one
example,
first
of
all
will
be
kind
of
nice
and
good,
and
second
one
we
say
like.
Oh
there's,
something
goes
wrong
with
ownership.
B
Ideally,
we
should
probably
kind
of
raise
air
over
there,
but
we
can
just
simply
swallow
it
and
just
delete
them
directly
was
like
today
on
environment
destruction.
We
just
we
just
deleted
objects,
we
don't
care
and
because
they
deleted
in
random
order.
We
can.
We
will
always
have
these
issues,
as
we
see
today,.
A
A
It
currently
makes
an
assumption
that
the
objects
do
not
rely
on
each
other,
I
guess
and
I
think
that
is
a
false
assumption,
at
least
in
this
case,
because
the
object
rap
relies
on
the
thread:
safe
function
to
be
there,
and
if
there
is
this
sort
of
dependent
internet
interdependency,
then
we
have
to
change
how
you
know
you
I,
don't
want
to
say,
like
the
order
of
things
being
destroyed
or
cleaned
up,
but
at
least
the
the
process
of
cleanup.
A
A
All
right,
that's
all
I
got
for
this
one,
all
right.
That
was
that
one
and
then
what
is
this
one?
Okay?
Yes,
this.
B
B
B
No,
if
the
API
is
table
node,
like
somebody
put.
A
B
And
I
think
we
need
to
kind
of
I,
don't
know
much
about
this
area,
but
would
be
nuts
if
we
have
it
so
now
later.
A
Oh,
this
is
a
relatively
new
issue.
Okay,
let's
see
essentially
calls
to
create
external
buffer
failed
with
array,
buffer,
new
backing
store
or
NBA
Town.
Everyone
must
be
out
of
here.
A
A
A
Yeah
I
think
we
can
take
take
a
look
at
the
I
think,
deprecating
or
discouraging
the
use
of
an
API
create
buffer,
create
external
offer.
B
So
should
we
kind
of
our
code
should
be
aware
with
what
mode
we
have
for
V8
like
if
it's
sandbox
mode?
Do
you
call
different
apis,
I,
don't
know
really
nice.
B
Oh,
we
still
use
node
buffer.
A
Yeah
I'm
not
not
too
sure
about
that
they
may
be
related,
but.
B
Yeah,
okay,
but
again,
I
think
instead
of
introducing
I,
would
rather
have
maybe
feature
which
will
just
simply
fix
kind
of
provide
a
fix
for
existing
API.
So
because.
A
B
A
A
Let's
put
that
in
okay,
that
is
it
for
still
issues
we
have
about
five
minutes
left.
Does
anybody
have
any
issues
that
they
would
like
to
discuss
in
any
repositories.
B
Nope
I
guess
I
only
have
if
no
no
other
kind
of
issues
are
I.
I
have
a
question
like
I've
seen
like
a
some
of
our
Republic.
B
We
have
a
little
bit
outdated,
npm
packages,
wonderful
if
it's
okay,
if
I,
can
update
them
to
the
latest
and
greatest
versions.
A
I,
don't
see
a
problem
with
that.
Yeah
I
also
noticed
that
in
the
like
the
example,
repo.
A
Did
I
make
that
up
I
did
make
it
up;
I,
guess:
no,
not
no!
No!
No!
No!
No!
What
I
saw
is
that
probably
what
you're
referring
to
yeah
like
they're,
using
older
within
here
and
then
I'm
guessing
you're,
also
meant
talking
about
within
our
package
of
node
add-on
itself.
Maybe
some
of
these.
A
B
A
I
I
don't
see
a
problem
with
that
at
all.
If
we
can
find
some
way
to
update
those
okay,
okay,
I
wanted
to
look
at
issues
that
are
new.
There
was
one
that
was
coming
in
I.
Think
Jack
has
been
working
with
this
person.
A
Yeah
I'm,
just
gonna,
wait
for
I've
been
sort
of
following
this
too.
Just
wait
until
this
person
comes
back,
then
that
is
that's
it
for
there.
A
That's
the
newest
one.
There
I
feel,
like
I,
saw
a
PR
recently.
A
Not
too
sure
what
that
is
supposed
to
to
modify
but
I
think
there
has
been
some
earlier
discussion
here.
A
A
I
think
since
maybe
if
Legend
the
cast
comes
back
to
this
issue,
I
mean
we
can
also
look
at
where
this
finalized
Rand
variable
is
used.
B
I
think
he's
looking
for
like,
maybe
maybe
we
need
to
have
an
extra
unit
tests.
If
we
have
such
addition
today,
it
would
be
nice
to
have
you
know
this.
B
It
would
be
ideal,
a
definitely
kind
of
fixed,
looks
trivial,
but
I
always
wonder
how
many
other
kind
of
things
like
this.
We
we're
going
to
have
and
because
we're
doing
active
changes
in
this
area
change
on
change
code
there,
and
we
need
to
make
sure
that,
after
his
change,
this
test
case
would
continue
to
work.
So
modest
case
is
better
right
right.
B
Because
otherwise,
we're
talking
about
like
yes,
we
have
some
application,
have
memory
leak
after
it's
fixed,
it
does
not
have
memory,
but
better
engineering
approach
would
be
to
have
a
test
which
shows
this
issue
and
after
the
fix
it
shows
it.
It's.
A
Actually
gone
right,
test
driven
development,
okay,
well,
I
think
that
that
is
it
for
the
new
issues,
or
at
least
things
that
I
have
seen
recently
yep
okay,
so
that
is
it.
We
don't
have
any
more
time
just
quickly.
A
There
was
this
last
issue
on
the
milestones
for
creating
documentation
on
possible
memory
leaks,
but
I,
don't
think
any
additional
work
has
been
done
on
this
all
right.
Well,
I!
Guess
that
is
it
for
this
week's
call?
Does
anybody
else
have
anything
they
would
like
to
quickly
discuss.