►
From YouTube: WEBRTC WG interim 2021-06-22 (partial)
Description
The last 10 minutes of the meeting (a pilot error prevented getting the rest).
A
Because
of
that,
it
could
not
provide
the
dfinity
a
comprehensive
list
of
all
of
the
like
if
you,
if
the
user
ends
up
choosing
this,
I
want
to
do
that.
If
the
user
chooses
that,
I
want
to
do
something
else.
I
think
that's
a
bit
of
a
difficult
thing
to
do
for
the
application,
so
that
and
the
application
will
not
find
out
what
the
user
selected
or
what
was
even
available
until
the
user
clicks.
Ok,.
A
So
it
seems
to
me,
like
the
only
difference,
would
be
that
we
do
it
like
up
to
one
second
later,
which
you
know
is
a
window
that
we
could
shrink,
and
I
don't
think
that
the
user
would
really
realize
if,
like
hey,
we
did
it
after
10,
milliseconds
or
250
beyond
that
they
might
notice
yes,
but
basically
I'm
giving
you
the
mechanism
by
which
the
decision
is
made
like
capture
handle
means.
Okay,
I
want
to
if
it's
this
url
or
this
url
or
if
the
handle
is.
B
A
Okay,
thank
you
for
the
feedback.
I
I
think
that
it's
difficult
for
me
to
imagine
what
kind
of
picker
there
would
be
where,
as
I
scroll
and
highlight
different
tabs,
suddenly
checkbox
would
kind
of
change
its
default
value.
And
then,
if
I
highlight
one
thing
suddenly
there
is
a
check
box
that
says
oh
that's
on,
and
then
I
don't
turn
it
off.
I
just
highlight
a
different
tab
and
suddenly
it's
on
and
you
know
it.
C
All
right,
so,
yes,
I
also
agree
with
tim
here
that
I
also
think
there's
some
security
concerns
and
that
you
know
we
don't
have
that.
We
have
these.
I
have
flashbacks
to
web
pages,
that
open
windows
and
behind
the
browser-
and
I
don't
notice
them
until
I
close
my
window
so,
but
I
think-
and
it's
true
that
we
do
push
all
the
user
agents
today
push
the
capture
surface
to
front,
which
I
think
is
evident
that
this
api
is
about
presentations.
I
mean
that
is
the
primary
use
case.
C
So
I
think
that
is
evidence
that
these
decisions
are
in
scope.
I
don't
like
the
api
here,
though,
because
hey
I
don't
think
application.
The
application
necessarily
knows
best.
It
might
depend
on
the
source.
That's
chosen
what's
going
to
happen,
but
also
they
seem
like
hacks
frankly,
in
that
it's
clear
that
get
display
media,
even
though
the
use
case
is
presentations.
C
That's
the
primary
use
case,
the
actual
tool
is
more
of
a
blowtorch
and
that
it
has
the
security
properties,
so
the
user
agent
is
is
not
fully
aware
of
the
context
here.
So
I
think
a
better
way
to
approach
this.
Would
be
to
to
elevate
these
concepts
of
safely
captured
pages
that
are
presentations
and
that
the
user
agent
knows
what's
going
on
it
can
make
the
best
decision
of
when
to
push
things
to
front
or
not.
A
Okay,
so
I
would
just
say
that
the
application
does
no
using
capture
handle
what
surface
was
chosen.
It
just
knows
it
immediately
as
we
resolve
the
promise
and
the
user
agent
can
use
the
very
same
thing
so
capture
handle
could
be
used
for
the
user
agent
to
make
a
default
decision
if
the
application
failed
to
do
that,
and
I
think
that,
yes,
even
if
we
say
that
get
display
media
is
mostly
for
presentations,
which
I
don't
say
it
is.
But
let's
accept
that
argument
for
the
sake
of
argument.
A
A
There
is
still
a
need
for
this
and
we
can
debate
exactly
how
to
achieve
this,
but
I
think
that-
and
I
I
consider
that
it
is
a
risky
proposition,
but
I
think
that
with
a
very
short
window
of
when
you
can
affect
the
change,
it's
not
so
risky
right,
it's
not
like
you
could
hold
it,
and
change
focus
five
seconds
later
or
five
minutes
later.
It
closes
after
one.
Second,.