►
From YouTube: 2022-04-06 Monthly Multicast W3C Community Group Meeting
Description
Notes here: https://docs.google.com/document/d/1MMvQuMfDtcj_cRVPWDnVUMNKWhcmF9um3WsUXK5EC3E/
Meeting Series: https://github.com/w3c/multicast-cg/tree/main/meetings
A
Great
recording
is
on
okay
hi.
Everyone
welcome
thanks
for
coming,
so
I
don't
have
too
big
an
agenda
today,
but
there's
there's
several
new
people
attending
so
I'd
be
happy
to
add.
Maybe
a
bit
of
introduction
if
anybody's
willing
to
kind
of
go
around
say
who
they
are
and
what
drew
their
interests
here
and
and
what
they
hope
to
do
going
forward.
Yeah.
B
Two
minutes:
yeah
go
ahead:
hi
everybody,
I'm
a
new
member
of
the
community,
so
I
live
in
paris.
I'm
a
fresh
student.
B
A
I
I
can
also
introduce
myself
to
those
of
you
who
haven't
seen
me.
I'm
I'm
jake
highland
I've
been
working
on
multicast.
We
I
I
think
I
haven't
talked
directly
yet
with
dom
before
I.
I
didn't
see
your
picture,
so
I
I
don't
remember
if
I
know
you
I'm
sorry
hold
on.
Let
me
I'm
in
a
scruffy
mood,
so
there
you,
you
don't
have.
D
E
Yeah
so
please,
let's
yeah,
I'm
going
to
turn
it
off,
though,
because
I'm
going
to,
I
can't
I
thought
I've
got
to
hold
my
arm
out.
D
All
the
time,
so
I
I
yeah
just,
I
suppose,
a
few
words
of
introduction
and
just
I
I
was
interested
very
much
in
multicast
in
the
early
2000s
built
a
cdn
with
a
chat
called
tony
balardi.
Who
was
one
of
the
sort
of
early
inceptors
of
amt
and
we
ended
up.
D
He
ended
up
very
focused
on
mbgp
as
a
central
part
of
the
networking
for
our
peering
strategy
at
london
internet
exchange
in
the
early
2000s,
and
we
were
focused
on
driving
all
our
live
traffic
across
access
networks
and
you
know
getting
igmp
out
to
the
access,
routers
and
blah
blah
blah
and
doing
all
that
sort
of
stuff.
It
was
great
fun
for
nine
years
and
then
we
we
couldn't
deal
with
the
onslaught
of
the
big
us
cdn's
arriving
in
in
the
uk.
So
I
switched
to
selling
streaming
software.
D
That's
what
I
do
today,
but
was
enthralled
through
a
sustainability
project
to
hear
jake's
name
and
made
a
connection,
and
hence
I'm
here.
So
I'm
curious
to
know
what
you
guys
are
up
to
and
where
multicasting
has
gone
over
the
last
decade.
Really.
F
Yeah
sure
wait:
does
my
camera
work
yeah
there
you
go
hi,
I'm
max!
I'm
a
now
a
phd
student
at
tu
berlin.
I
first
met
jake
like
three
years
ago
now
at
itf
through
tubs,
which
he
tried
to
use
as
a
way
to
deliver
multicast
that
didn't
work
out
so
good.
I
guess,
but
it
was.
F
It
was
a
good
thing
because
I
met
him
and
implemented
army
one
of
the
the
protocols
or
the
drafts
he
has
and
now
I'm
just
trying
to
focus
on
multicast
during
my
phd
studies
right
so
any
anything
related
to
that.
I'm
looking
into
that
and
I'm
also
hoping
to
like
bring
it
more
to
the
students
and
stuff.
So
we
hope
to
have
like
a
student
project
next
semester
about
multicast
and
also
make
it
maybe
part
of
our
lectures
and
stuff
so
yeah.
That's
that's
what
I'm
working
on.
G
Kyle,
I'm
kyle
rose,
I'm
an
architect
at
akamai.
G
I
work
somewhat
with
jake,
I'm
also
the
one
of
the
co-chairs
of
the
media
operations
working
group
at
ietf,
and
I
have
a
my
my
job
since,
since
I
moved
organizations
about
a
year
or
so
ago
is
is
to
you
know
like
essentially
be
a
an
architect
for
forced
strategic
initiatives
around
the
company
and
one
of
them
at
I.
I
work
at
akamai
with
jake,
as
I
said,
and
one
of
them
is
is
multicast,
so
I've
been
trying
to
spend
time
on
this.
I
Chris
needham
sure
yeah,
so
hi,
I'm
chris.
I
work
for
the
bbc
in
the
uk.
We
are
certainly
interested
in
multicast
distribution.
You
know
for
for
live
video
content,
I'm
not
particularly
a
multicast
specialist
myself,
I'm
mostly
active
at
w3c
in
in
web
standards
there.
I
So
I
I'm
a
co-chair
of
a
interest
group
which
looks
at
new
new
sort
of
media
related
applications
on
the
web
and
also
a
co-chair
of
the
media
working
group,
which
is
standardizing
a
bunch
of
apis
for
media
streaming
and
related
things,
things
like
media
source,
extensions
and
crypto
media
extensions,
media
capabilities
and
so
on
and
kind
of
met
jake
through
his
his
work
on
building
a
chrome
prototype
implementation.
A
Yeah
chris
is
chris's,
help
was
instrumental
in
getting
this.
This
community
group
kicked
off
so
yeah,
thanks
for
that
and
and
all
his
great
work
here
at
w3c
yeah
chris
leonard.
J
You're
willing
yeah,
I'm
here,
okay,
so
yeah,
I'm
chris
leonard.
I
work
for
verizon
been
talking
to
jake
for
a
few
months
now,
maybe
going
back
to
middle
of
last
year,
but
I,
I
guess,
sort
of
kind
of
I'm
the
multicast
guy
within
verizon.
J
And
you
know
do
what
I
can
to
help
as
far
as
discussing
it,
at
least
and
and
then
also
just
you
know,
kind
of
participating
on
the
sidelines
to
see
what's
coming
up
and
what
might
be
useful
in
my
company,
and
you
know
talking
within
the
industry
with
other
companies
and
other
participants
and
standards
bodies
to
kind
of
you
keep
it
moving
forward.
A
Great
yes,
thank
you
very
much
thanks
everyone
for
being
here
and
for
your
intros.
That
was
great,
so
I
just
had
a
few
things
that
were
kind
of
I
put
down
as
being.
A
What's
on
my
mind
this
this
session,
and
I
guess
the
the
first
of
these
I
had
down
as
the
just
I
started
off
on
the
multicast
quick
trying
to
poke
into
the
code
a
bit
and
figure
out
what's
going
on,
I
feel
like
I
haven't
gotten
very
far,
yet
you
know
I'm
getting
my
head
around
the
code
there.
Anybody
has,
you
know,
advice
on
alternatives
on
that,
I'm
very
open
to
suggestions
there,
but
you
know
so
far.
A
I've
just
you
know,
found
the
place
where
pvp
sockets
are
opened.
I'm
going
to
have
to
put
something
in
there,
because
when
you're
sending
sending
multicast
you
have
to
set
the
ttl
and
when
you're
receiving
multicast
you're
going
to
have
to
do
a
stock
cop
for
it.
A
So
you
know,
and
what
else
I'm
I'm
trying
to
figure
out
like
where
the
right
places
to
put
the
frame,
there's
gonna,
be
frames,
for
you
know
negotiating
support
for
this
for
advising
a
receiver
to
join
a
particular
group,
as
well
as
the
the
sort
of
multi-path
based
like
packing
of
a
different
packet
number
space.
I
wish
I
could
find
the
the
I
kind
of
wish
the
multi-path
work
was
a
little
further
along
and
published.
A
I
haven't
been
able
to
find
an
implementation,
yet
that
is
the
that
is
the
one
that
they're
that
they're
currently
trying
to
work
on.
So
if
anybody
knows
about
that
stuff,
then
yeah
send
it
my
way
or
if
you
encounter
it
but
outside
that,
I'm
you
know
just
starting
to
to
plug
along
on
that.
Let's
see
any,
I
guess
comments
or
questions
on
that
point.
A
This
last
deed
had
some
distraction
on
the
you
know
from
from
doing
the
the
sort
of
more
important
work
of
moving
the
deployment
conversations
forward.
Those
also
are
slow.
It's
not
like.
I
have
things
that
are
public
to
say
about
it,
but
but
they
are
moving,
and
it
is
very,
very
important
to
this
project
that
they
do
move
so
so
that
interrupts
this,
this
quick
work
whenever
it
happens,
but
other
than
that
I
I
am
trying
to
get
the
quick
working,
because
that
is
the
way
forward
in
the
browser.
D
A
No,
that's
actually.
I
I
really
should
have
included
the
link
to
the
proto
draft
that
we're
gonna
that
we're
gonna
be
pulling
forward,
which
I
will
include
now
in
the
notes
or
in
the
so
here's
the
link
in
the
chat.
There
are
a
few
paragraphs
in
there,
so
just
a
little
background.
The
the
multicast
work
that
I've
been
doing
is
about
is
about
ingesting
multicast
from
an
external
source.
A
So
if
you
look
at
rfc
8777,
this
is
the
the
dryad
spec,
which
is,
of
course,
the
built
acronym
out
of
dns
reverse
ip
for
amt
relay
discovery.
I
think
something
like
this,
so
the
the
idea
is
that
when
there
is
a
ssm
join
given
to
a
network,
the
the
goal
is
to
end
up
with
with
multicast
ip
over
the
last
mile
into
homes.
A
Okay,
because
that
is
where
the
big
pain
point
is
that
that
can't
really
be
solved
in
another
way
and
so
the
that
last
mile
you
know
this
is
like
in
your
in
your
gpon
fanout
kinds
of
deployments
or
in
your
cable
loop
deployments,
you
can.
A
You
can
get
multicast
data
to
be
shared
between
all
the
local
receivers.
In
a
loop
right,
so
today,
where
you've
got,
you
know
your
10
megabit,
video
stream,
when,
when
you're
watching
you
know
the
world
cup
or
something
and
there's
a
10,
megabit
video
stream,
let's
say
80
of
the
homes
in
an
isp,
then
each
one
of
those
homes
is
getting
this
unicast
stream
delivered
to
them.
A
And
you
know
when
you,
if
you
can
do
that
by
multicast
over
that
last
mile,
then
it
the
entire
space
of
that
can
be
shared,
like
all
the
data
that
can
be
transmitted
to
everyone
can
be,
can
just
take
up
that
110
megabit
stream
on
on
the
physical
connection,
and
so
that's
the
goal
that
we
want
to
realize.
With
this.
A
Quick
or
a
native
multi,
all
right,
it
would
be
native
ip
multicast
and
the
idea
would
be
so.
The
idea
with
quick.
You
know
the
there
there's
another
draft
that
that
was
from
the
bbc
about
about
http
over
quick
multicast.
A
A
But
the
idea
is
to
do
something
like
multi-path,
where,
where
there
is
a
unicast
anchor,
but
that
in
that
unicast
anchor
stream,
the
server
will
tell
the
client
about
alternate
sources
of
data
and
those
sources
of
data
being
a
multicast
ip
stream
that
the
client
can
put
it
in
yeah.
A
And
then
you
know
the
so.
The
the
anchor
unicast
stream
will
be
used
for
all
the
security.
For
you
know
all
the
security
guarantees
that
apply
to
the
to
the
data
received
on
the
other
layer,
but
you're
going
to
try
and
get
all
the
data
that
you
can
into
that
into
that
multi-cast
stream
or
all
the
data
into
that
multicast.
A
So
that's
that's
the
basic
idea
and
that's
that's.
What
we're
trying
to
do?
Yeah
it's
it's
gonna,
be
a
bit
of
a
road.
I
don't
know.
I
guess
you
probably
haven't
been
following.
We
had
a
if
you,
if
you
follow
that
link
and
read
through
it,
you'll
find
the
the
security
considerations
document
that
we
presented
in
november
at
ietf
at
this
at
the
sec
dispatch
working
group
to
see.
If
we
could
get
some
discussion
kicked
off.
A
You
know,
I
guess
my
summary
interpretation
is
that
most
of
the
web
security
people
are
basically
not
very
interested
in
multicast.
They
they,
I
think,
what
they
you
know.
This
is
my
interpretation,
but
I
think
they
do
not
think
that
it's
plausible
that
it
will
end
up
working
and
therefore
it's
not
worth
spending
their
time
on.
A
I
think
people
who
are
more
sort
of
performance
oriented
tend
to
see
the
value
in
in
getting
multicast
to
work,
that's
kind
of
where,
where
I'm
coming
from.
In
my
background,
but
you
know
there
are
some
security
challenges
and
there
are
some
fundamental
security
challenges
that
have
to
do
with
the
sharing
of
the
data,
especially
in
terms
of
privacy,
is
where
people
have
kind
of
expressed.
A
The
most
hesitancy
we
tried
to
capture
that
discussion
in
the
that's,
the
draft
k,
rose,
multicast
security
and
I
think
I
think
that
does
a
pretty
good
job
of
of
covering
the
issues
and
and
giving
our
take
on
it.
I
think
the
one
review
we
got
from
ecker
was
did
not
really
highlight
a
lot
in
the
way
of
technical
objections.
I
don't
know
if
you
agree
or
disagree
with
that
kyle.
I
think
we
had
a
couple
of
like
oh
yeah.
We
should
add
that
point.
A
A
But
on
the
on
the
flip
side
of
that
is
that,
in
order
for
this
to
to
really
meet
the
needs
of
the
internet
in
the
long
term,
it's
going
to
have
to
support
web
traffic.
Somehow
you
know
to
to
sort
of
give
the
incentive
for
isps
to
to
do
that
work
of
rolling
it
out.
You
got
any
opinions
on
that
chris
because,
because
I
think,
that's
kind
of
my
summary
take
on
on
what
what
the
isps
are
going
to
think
of
this.
C
A
Not
not
necessary,
I
think
yeah,
if
you,
if
you
disagree,
dude,
please
chime
in,
but
but
that's
my
my
basic
position
there
right.
So
so
the
sorry.
A
It's
a
good
question,
absolutely
and
and
yeah
take
a
take
a
look
at
those
drafts
and
see
if
there's,
if
there's
anything,.
A
Yes,
there
is
all
right,
so
I
guess
any
other
questions
or
comments
on
that
part.
A
I
I
don't,
I
think
that
so
I
did
talk
to
lucas
and
he
was
he
was
gonna.
Join
me
on
this
on
this
quick
one
spoke
a
little
bit
with
with
the
other
authors
as
well
of
the
http
multicast.
They
had
been
keeping
it
alive
for
some
time.
I
think
you
know
I
don't.
I
don't
have
like
a
a
really
good,
solid
answer
from
from
the
other
authors
on
the
http
quick
multicast.
So
I
don't
really
know
the
intention.
A
A
The
removal
of
hdb
push
from
the
browsers
getting
this
into
web
transport.
You
know
nobody
has
disagreed
yet
that
that's
a
that's
a
good
approach
to
getting
this
done.
You
know,
I
think,
one
day
if
we
do
get.
If
we
do
get
this
deployed,
then
maybe
push
will
come
back
because
I
do
think
it's
going
to
be
useful,
but
it
will
probably
have
to
be
done
in
the
web
app
layer
for
a
while
through
web
transport
because
they
don't
just
take
out
push
and
then
put
it
back
in
another
year
later.
A
So
yeah
any
did
that
answer
your
question,
so
I
I
don't
know
the
status.
I
guess
is
my
answer.
Thanks.
A
A
F
F
A
A
Yeah
thanks
yeah
matt
max
noticed
in
some
of
his
rust
work
that
that
the
the
current
rust
socket
api
only
has
support
for
asm
multicast,
so
his
he's
looking
at
adding
ssm
multicast
as
well,
and
that
would
that
would
make
it
more
reasonably
possible
to
to
port
into
like
the
firefox
implementation.
A
I
think
so.
That's
certainly
a
worthwhile
thing
to
add,
but
I
think
it'll
be
more
work
as
well.
We'd
want
to
get
the
quick.
A
The
quick
running
somewhere
yeah,
I'm
not
the
only
reason
I
I
look
at
chrome,
so
hard-
is
just
because
they're
deployment
footprint,
but
I
think
neither
firefox
nor
nor
chrome
is
particularly
interested,
so
it
doesn't
really
matter
which
one
it's
not
like
doing
it
in
one
or
the
other
makes
it
more
likely
to
deploy.
A
We're
gonna
have
to
get
the
the
ball
rolling
through,
demonstrating
that
isps
are
willing
to
do
the
forwarding
first.
I
think.
J
J
That's
all
that
definitely
helps
me
is
as
far
as
moving
forward
with
what
I
need
to
do,
and
then
you
know,
as
you
said,
having
the
capability
of
doing
web
traffic
is,
is
probably
to
be
a
big
step
to
getting
the
best
coverage
as
well.
A
Yep
yep
yeah,
there
are
other
important
platforms,
especially
with
the
software
delivery
kind
of
angle,
but
but
for
video
I
think
I
think
web
is
is
really
critical,
especially
when
you
count
that
a
lot
of
smart
tvs,
really
what
they're
doing
inside
is
that
with
access
to
the
to
the
web
browser
javascript
apis
and
such
so.
That's
like
a
lot
of
times.
A
Those
apps
you
see
on
the
smart
tvs
are,
you
know
an
implementation
of
a
of
essentially
what's
a
like
a
web
app
in
a
constrained
environment,
kind
of
a
thing
not
strictly,
but
I
think
that's
the
that's
the
way
the
wind
seems
like
it's
blowing
in
that
space.
As
far
as
I
can
tell.
A
And
one
day,
they'll
probably
start
updating
those
more
regularly
too
just
because
of
all
the
security
issues,
but
the
ones
that
are
running
now
apparently
are
somewhat
behind
and
don't
auto
update,
but
or
at
least
don't
auto,
update
very
frequently.
But.
I
H
I
A
A
What
else
did
we
have?
So
I
mean
you
expressed
an
interest
in
the
amt
work.
I
I
threw
that
in
there
it
was.
It
was
my
work
at
the
at
the
ietf
hackathon.
I
guess
that
was
what
two
weeks
ago.
A
Did
you
have
any
questions
about
that
because
the
we
do
use
it
in
a
lot
of
the
sort
of
here's
how
you
set
it
up
so
that
you
can
pull
external
traffic
kinds
of
demos.
D
It's
just
it
was
really
interesting.
As
you
said
later,
the
papers-
and
I
was
trying
to
you-
know,
get
tuned
into
what
the
group
was
focused
on
seeing
amt
there
and
tony,
and
I
had
a
tony
blighty
and
I
had
a
very
close
relationship
with
amt
trying
to
work
out
how
to
get
around
the
art
of
the
access
routers,
not
having
ignpv3
available,
and
so
we
sort
of
built
a
tunnel,
an
auto
tunneling
system,
which
then
would
relay
relay
multicast
into
a
lan
basically
and
have
a
nomination
signaling
system.
D
It
was
all
kind
of
something
you
know
the
variance
of
it
would
look
like
octa
shape
and
other
technologies
like
that,
a
little
bit
later
so
yeah
I
had
a
real,
close,
close
interest
in
it.
We
were
trying
to
deliver
it
as
a
browser
plug-in.
So
it's
really
interesting
hearing
what
you're
just
talking
about
five
minutes
ago
and
thinking
that
was
2006.
We
were
trying
to
do
that
so
a
little
bit
window
windows,
media
being
the
baseline.
At
that
point,.
A
D
It
was
a
pragmatic,
it
was
a
pragmatic
solution
because
we
we
as
fast
as
we
got
isps
to
appear
with
us
in
in
in
the
exchange
and
manage
their
routers
to
do
the
igmp,
v2
and
v3
forwarding
some
cto
would
change
and
come
in
and
go.
I
don't
know
what
that
is.
It's
not
vanilla
enough
turn
it
off
so
yeah.
A
Yeah
the
take
the
the
approach
we're
taking
now
is
actually
the
the
main
reason
I'm
interested
in
amt
is
to
run
it
at
the
ingest
of
the
isp,
so
the
idea
is
that
you've
gone
and
made
it
again.
You
might
be
interested
in
another
link
that
I
didn't
include
this
time,
but
the
multicast
receiver
platform.
A
This
is
what
we've
trialled
with
with
a
bunch
of
isps
that
have
been
evaluating
our
stuff
and
this.
The
idea
is
that
you
would
you
put
the
multicast
ingest
platform
in
in
your
in
your
network,
where
you
want
it
to
be
multicast
native
multicast
from
their
down
toward
end
users
right.
So
I
have
a
couple
of
these
in
my
house,
south
of
my
of
my
wi-fi
home
router,
but
the
other.
A
You
know
the
other
main
place
to
run.
It
is
like
in
a
real
isp,
where
you
have
a
real
multicast
network
that
goes
from
there.
So
the
idea
is
that
that
you
deploy
these
and
then
it's
sort
of.
Like
a
I
mean,
I
think,
of
it
in
terms
of
more
like
a
http
or
multicam
in
a
sense
where
it's
just
you
don't
worry
about
what
the
stream
is.
You
don't
you
know
you
don't
have
to
care
where
it
came
from.
A
To
to
pull
multicast
in
as
long
as
the
sender
has
set
up
the
sort
of
right
infrastructure
to
advertise
that
can
be
discovered
by
that
ingest
platform.
D
A
D
Explore
beer.
A
Yes,
so
I'm
certainly
aware
of
beer-
and
I'm
I'm
looking
at
it
to
me,
beer
is-
is
great
for
the
isp.
That's
doing
the
ingest
is
kind
of
the
way
I
look
at
that
they
they.
It
really.
A
Of
that
multicast
network
of
the
multicast
capable
network,
I
think
of
it
as
a
walled
garden.
There
are
some
efforts
to
open
that
up
so
that
you
can
have
like
the
ifrs
and
the
befrs
not
part
of
the
same
domain,
but
I'm
not
sure
how
how
that's
going
to
go
exactly.
I
don't
know
what
the
trust
model
on
that
looks
like
really
yeah
it's
it's.
I
mean
that
is
possible,
but
it's
like
another
kind
of
peering.
I
think
which
is
anyway
yeah
beer.
A
What
beer
really
buys
you
is
that
you
don't
need
the
hop
by
hop
state
when
you're
doing
the
when
you're
doing
the
in
the
network.
That's
doing
the
forwarding
so
yeah
we
we
do
mention
beer
to
the
isps
we're
talking
to
and
and
sort
of.
A
I
think
it's.
I
think
it's
going
to
be
really
helpful
for
the
future
of
multicast
and
how
things
get
deployed.
But
that
to
me
is,
like
you
know,
as
akamai
providing
like
traffic,
you
know
we
stand
up
a
way
to
ingest
traffic
and
then
an
isp
will
ingest
it
and
then
what
they
do
with
it
is
kind
of
up
to
them
up
to
them.
A
Yeah,
you
know,
but
beer
is
a,
is
a
very
good
way
to
do
the
delivery,
in
my
opinion,
so
I
think
it's
going
to
be
like
a
lot
of
the
isps
that
look
at
this
will
end
up
doing
that,
but
but
outside
that
we
you
know,
I
haven't
dug
into
beard
very
deeply.
I'm
just
like
any
multicast
support
would
be
great,
but.
D
Yeah,
I
think
greg
shepard
would
probably
be
good
to
get
fired
up,
he's
always
a
fun
for
this.
A
Yeah,
I
mean
absolutely,
I
think
I
think
beer
does
provide
really
good
value,
but
you
know
and
but
to
me
whether,
like
on
that
last
mile,
where
it's
critical
to
do
the
to
do
the
multicasting,
it
doesn't
really
matter
whether
it's
beer
and
those
those
home
gateways
are
like
bfers
or
whether
it's
multicast
and
the
home
gateways
are
just
igmp
proxies
or
something
mld
proxies
so
but
yeah
either
way
is,
is
fine
as
far
as
I'm
concerned-
and
I
mean
yeah
beer
is
great,
but
but
not
critical
to
the
question
of
inter-domain
multicast.
D
A
What
else
do
we
have
on
the
agenda?
Any
any
questions
on
that
front,
the
amt
or
oh,
we
had
a
question
about.
Should
we
update
our
getting
started
doc,
I'm
not
sure
anybody's,
actually
read
it
or
used
it
yet.
A
Actually,
my
my
browser
releases
are
a
little
behind,
but
here
that's
under
primers
here
I'll
add
the
the
link
to
the
chat
yeah
don.
If
you
want
to
take
a
look
and
give
us
feedback
on
on
the
getting
started
page
and
especially
so
the
the
question
here,
because
the
amt
gateway.
A
That
we
put
in
a
couple
weeks
ago
mean
that
we
can
now.
We
could
now
include
some
some
instructions
for
ipv6
that
might
be
worthwhile.
I
have
an
ipv6
sender
running
that
can
be
that
can
be
used
for
for
pulling
some
of
that
traffic,
but
you
know
I'm
not
sure
how
critical
it
is,
especially
with
kind
of,
but
we
do.
I
think
it
is
important
to
have
good
entry
points
for
people
coming
into
this,
and
I
think
that's
one
of
them.
We
probably
also
max.
A
Do
you
think
you're
you
sent
me
the
link,
and
I
said
I
haven't
yet
you
had
an
intro
to
multicast
document.
I
think
we
should
perhaps
add
that,
to
our
add
a
link
to
that
in
our
primers.
F
A
pass,
I
guess
but
yeah
I
I
we
can
edit
sure
if
you
want
to
add
the
pdf
version
or
the
I
don't
know
which
one
because
I
think
the
link
to
overleaf
might
not
be
the
best
way
to
do
it.
Perhaps
yeah.
A
Let's,
let's
take
that
as
a
follow-up
action
item,
then,
if
you'll,
if,
when
you,
I
think
anybody
here
that's
willing
and
able
to
review
it,
please
please
take
a
look
but
max
if
you
could
send
a
link
to
that
to
the
list
and
then
we
should
you
know
unless
anybody
objects,
I
think
we
should
add
it
to
the
primers
list
for
how
people
can
get
oriented
with
understanding
and
grappling
with
the
problems
with
multicast
and
multicast
to
the
browser
in
particular,
but
I
think
it
so
max
started.
Writing
this.
A
It's
it's
intended
as
an
introductory
document
to
kind
of
get
people
oriented.
So
I
think
it
would
be
a
really
good
addition
to
our
to
the
community
groups
list
of
primers
that
currently
contains
one
entry
that
chris
chris
needham
and
I
worked
on
a
few
months
back,
but
I
think,
having
some
more
content
in
there
to
help
people
get
started
would
be
really
good.
A
D
A
I
don't
know
it's,
that's
fine,
that's
fine,
but
sure.
G
A
Interesting
webex
or
perhaps
webrtc,
I
I'm
not
using
the
web
interface,
I'm
using
the
the
app
that's
interesting
all
right.
Let's
see
anything
else,
we
should
go
over.
A
So
that's
kind
of
my
my
intended
activity
for
the
next
for
the
next
month.
On
this
front.
A
So
I
guess
anyone
else
want
to
volunteer
for
for
activity.
I
don't
know
if
I
don't
know
people's
level
of
interest
in
in
trying
to
participate
in
that
or
get
or
do
a.
I
don't
know
either
testing
or
digging
into
the
code
with
me,
we
could
do
a
working
session.
We've
done
a
couple
of
those
before
if
anybody's
interested
I'd
be
happy
to
set
one
up,
otherwise
I'll
just
continue
kind
of
hacking.
G
I
mean
I
would
be
interested.
I
just
I
want
to
make
sure
that
that
it
it
winds
up
being
helpful
for
you
not
just
a
demo
for
me
right.
You
know
what
I
mean
yeah.
A
Okay,
well,
some
you
know
you've
been
following
quick
longer
than
I
have.
You
might
have
good
advice
on
on
what's
going
on,
but
I
don't
know
I
don't
know
if
anybody's
dug
into
the
quick
code
base.
Yet
that's
here
not
really.
F
I
don't
know
I
I
would
also
be
interested,
but
for
me
like
the
hurdle,
is,
I
guess,
the
chromium.
I've
never
worked
with
the
chromium
source
before
so
right
yeah.
But
I
I,
as
I
told
you
last
week,
I
am
read
the
quick
draft
and
I
took
a
lot
of
notes
where
I
saw
like
conflicts
with
multicast,
so
I
am
happy
to
contribute
that
way,
at
least
but
yeah.
I
would
be
totally
on
board
with
testing
it
or
helping
in
any
way.
That's
actually
helpful
to
you,
as
kai
said,
cool.
A
Yeah
well
so
I'm
less,
I'm
less
troubled
with
with
whether
you
know
if
I,
if
I
spend
an
hour
or
six
on
on
a
working
session
and
the
one
day
isn't
helpful
to
me,
but
it
gets
you
guys,
engaged
and
and
lands
on
something
where
you
can
contribute
at
the
code
level.
Then
I'd
be
totally
open
to
that.
You
know.
I
think
that
would
be.
I
mean
I
I
kind
of
agree.
A
I
don't
I
don't
want
to
just
like
do
a
demo
for
you
and
that's
the
end
of
it.
What
I'd
like
is
is
more
like.
We
end
up
in
a
in
a
good
place
in
terms
of
like
things
where
we
have.
I
we
have
a
concrete
path
forward
and
like
things
that
we
could
each
contribute
to
the
thing,
and
it's
not
just
just
an
abstract.
G
I
mean
so
so
I
mean
honestly
the
I
am
happy
to
help.
I
have
you
know
like
more
free
time
than
I
did
two
years
ago
right
I
mean
when
I
say
free,
I
mean
uncommitted
time.
I
have
more
discretion
over
what
I'm,
what
I
spend
my
time
on,
I'm
relying
on
you,
and
maybe
this
is
wrong,
but
this
is
what
I'm
doing,
I'm
relying
on
you
to
drive
it.
So
I
kind
of
in
some
sense
want
you
to
tell
me
what
to
do.
A
But
I
will
look
for
good
ways
to
do
that.
Certainly,
I
think
max
I've
I've,
given
you
some
some
ideas
on
direction
that
are
not
really
about
the
quick,
but
not
really
about
the
quick
and
browser
stuff.
F
A
Would
say
actually
getting
active
use
more
quickly
is
more
important
than
the
quick
work
like
I
said,
that's
that
is
that
is
superseding
my
quick
work
within
akamai
when
I,
when
I
get
a
deployment
when
I
have
it,
when
I
have
the
opportunity
to
spend
time
meaningfully
moving
forward
the
deployment
of
active
use
of
real
multicast,
not
just
test
multitask
right,
then
that
makes
a
bigger
difference.
So
to
me,
the
other
stuff
we've
talked
about
is
probably
a
better
investment
of
time.
A
A
So
if
you
think
that
that
will
get
you
to
the
right
place,
then
that's
probably
the
first
place
to
spend
to
spend
time
when
you
can
and
then
as
a
second
priority
like
whenever
there's
like
blocks
or
waiting
or
whatever,
then
that's
kind
of
where
I'm
putting
my
time
on.
That's
where
I'm
putting
my
time
on
the
quick
because
getting
to
where
we
have
a
viable
proposal
for
bra
is
probably
the
second
priority
to
me.
A
So
let's
see.
B
Excuse
me,
I
wanted
to
ask
in
which
language
do
you
want
us
to
program
the
the
extension
in
chromium
c
plus,
plus
or
arrest,
because
I
don't
have
a
lot
of
knowledge
in
euros,
but
maybe
I
can
can
learn
it.
A
For
chromium
it
would
be
c
plus
plus
for
for
firefox.
I
I
think
it
would
probably
be
rust.
Okay
yeah,
so
the
the
the
chromium
build
and
the
place
that
I'm
that
I'm
starting
that
I've
been
looking
is
the
c
plus
is
the
c
plus
side.
I
also
have
not
yet
used
rust.
It's
one
of
those
things.
That's
on
my
list,
but
I
haven't
got
there
yet.
A
I
already
never
want
to
use
c
plus
again,
but
but
it's
I
mean
languages
are
I
mean
I
actually
I
I
shouldn't
say
I
never
want
to
use
it
again.
It's
better
than
javascript
come
on.
A
But
we
also
want
to
be
on
europe
time,
so
I
I
will
set
one
up
for
let's
say
in
in
about
two
weeks:
if
that's
all
right
or
should
we
do
one
week,
I
gotta
check
my
account
well.
G
Yeah
see
what
the
see,
what
what
people's
maybe
send
out
a
doodle
like
figure
out
how
long
we
want
to
spend,
send
out
a
doodle
with
some
time
slots
and
we'll
see
what
works
out.
I
mean
I'm
happy
to
get
up
early
for
something
like
this.
I
do
my
best
software
development
in
the
early
morning
anyway.
G
A
All
right
all
right,
yeah
I
will
I
will.
G
A
That
that
makes
sense
you
know,
but
sometimes
the
hangover
also
can
be
an
impact.
So
I
need
to
make
sure
to
sleep
in
long
enough.
A
Yeah,
just
coffee
will
take
care
of
that
no
problem
anyway,
yeah
yeah,
so
I
I
will
aim
to
put
something
on
the
calendar
in
a
couple
of
weeks,
I'll
send
out
a
doodle
poll,
I'm
kind
of
thinking
four
hours
we'll
take
breaks
along
the
way
and
we'll
we'll
be.
You
know
we'll
kick
it
off
with
the
at
the
time
we'll
plan
to
spend
about
four
hours,
hacking
on
it
and
we'll
hope
to
end
up
with
with
something
reasonable.
At
the
end,
that
sounds
great.
F
D
F
B
A
Okay,
okay,
I'll
I'll,
take
that
into
account
as
well.
I
I
also
am
you
know
pretty
good
in
the
morning.
Jake.
A
You
know
it
really
depends
like
this
whole
concept
of
when
sleeping
begins
and
ends
is
really
just
a
you
know.
It's
not
really.
A
I
found
that
on
my
home
planet
days
appear
to
be
26
and
three-quarters
hours
long,
it
was
was
what
I
measured
when
I
when
I
just
tried
it,
but.
A
A
Well,
it
wasn't
really
starved
of
sunlight.
It
was
just
seeing
what
my
natural
rhythm
was
go
to
sleep
and
I'm
tired
and
wake
up
when
I'm
ready
to
wake
up,
and
the
answer
was
26
and
three
quarters
hours.
I've
never
felt
better
before
or
since
this
was.
This
was
for
basically
a
summer
in
college
and
and
it
was,
it
was
wonderful,
but
I've
never
been
able
to
pull
it
off
again.
A
So
anyway,
that's
the
minor
note,
but
but
this
is
why
my
schedule
varies
around,
as,
as
some
of
you
may
have
seen
anyway
great
then
I
I
will
set
up
the
at
the
invite
I'll
I'll,
send
it
to
the
list
and
see
if
anybody
else
wants
to
participate.
Also,
I
don't
know
if,
like
gavin
henry
has
been
doing
some
some
work
as
well.
A
A
I
don't
know
if
we'll
end
up
doing
that,
because
the
the
the
udp
receive
path
in
the
quiche
implementation
is
probably
where
we
have
to
end
up
being
to
do
the
quick
part.
So
it
might
be
well
we'll
have
to
see,
but.
A
A
Okay,
then
I
think,
if
that's
it,
then
then
that's
it.
Anyone
have
anything.
A
Thanks
very
much
chris
for
taking
notes
thanks
everyone
for
being
here
and
and
yeah
looking
forward
to
a
working
session,
good
to
good,
to
see
you
all
and
recording.