►
From YouTube: DASH Workgroup Community Meeting Sep 7 2022
Description
PR Review:
201 sync to main later today
202 wait until 201 is done before merging, but it is ready - Volodymyr gave it a try
203 ready to go
210 merge main back into 210 for fixes
211 side effects of Ixia C container IPv6 would be re-enabled. Also add a new feature to Ixia-C team to have a fix (at Keysight).
Migrating to ACR - Chris to document
A
Hi
everyone
thanks
for
coming
to
september,
7th
dash
community
call,
and
so
today
I
was
hoping
unless
someone
has
an
atom
they
want
to
cover.
I
was
hoping
we
could
look
at
prs
today
in
the
queue
unless
anyone
speak.
Now,
if
you
have
something
you'd
like
to
cover
specifically.
A
So
I
hope
everyone
can
see
I'll
go
to
the
prs
here
and
I
was
hoping
we
could
just
start
maybe
down
here.
I
need
to
pull
up
teams
and
see
if
jay
is
on
the
phone
here.
A
Okay
thanks,
so
then,
how
about
should
we
start
here
at
the
bottom
marcia
or
does
anyone
have
something
they
want
to
pick
out
of
the
queue
here.
C
Three
to
talk
about-
let's
talk
about
this
one
first
201,
because
I'd
left
some
comments
from
marion.
Maybe
if
he
wants
to
respond,
did
you
see
my
comments
about
sinking
to
maine?
No,
not
yet.
C
Okay,
the
pr
looks
good
and
I
I
cloned
your
branch
and
ran
it.
So
it's
really
nice
how
you
got
all
the
auto
generation
working
and
everything,
but
I
think
before
we
merge
this
to
maine.
We
need
to
you
need
to
sync
it
because
23
commits
behind
and
all
the
test
cases
and
other
things
are
going
to
be
impacted
by
these
changes,
and
I
think
if
we
don't
sync
it
and
make
sure
it
runs
in
the
ci
pipeline,
we
might
get
a
surprise
from
your
merchant.
C
C
And
then
that
one
there
a
couple
people
have
run
it
and
tried
it,
and
I
really
appreciate
the
feedback,
so
that's
ready
to
go,
but
I
want
to
wait
until
201
is
done.
So
we
do
this
kind
of
an
order
and
we
don't
have
you
know
any
other
kinds
of
surprises.
C
I
think
it's
kind
of
orthogonal
to
201,
but
still
doing
things
in
order
is
usually
a
little
bit
better.
So
pending
marion,
if
you
do
do
get
that
reconciled
and
it
runs
in
ci,
I
would
propose
we'd,
merge,
201
and
then
I'll
bring
my
other
pending
ones
up
to
snuff.
You
know
sync
everything
and
then
request
those
get
merged.
So
202
is
the
whole
docker
and
make
file
permission
and
user
identity
fixes.
C
203
has
also
been
looked
at
by
a
couple
people
and
I
did
respond
and
fix
a
couple
of
things
in
there
and
I
think
that's
ready
to
go.
It's
really
documentation
plus
one
make
file,
that's
a
reference
make
file,
it's
not
executed
and
that's
the
document
dash
as
a
sub
module.
C
So
a
couple
of
couple
people
have
kindly
looked
at
that
and
I
think
that's
ready
to
go.
I
did
have
to
fix
a
couple
of
typos
in
the
pensando
h,
a
proposal
in
order
to
get
the
ci
to
pass
and
update
the
word
list,
etc.
So
in
203,
if
we
merge
that
it's
also
going
to
change
some
change,
the
h
a
proposal
with
a
couple
of
typos
we'll
fix
those.
It
wasn't
really
my
intention
to
do
that.
But
I
didn't
want
that
ci
to
have
any
kind
of
red
x
on
it.
A
A
A
C
A
C
Well,
thanks
and
yeah
211
fix
ipv6
packet
noise.
What
that
is
about
is
due
to
some
side
effects
of
launching
xcsc
container.
It
would
re-enable
ipv6
after
I
disabled
it
in
the
linux
networking
to
do
software
packet
time,
testing
and
then
you'd
get
linux.
Ipv6
control,
plain
junk
would
start
hitting
the
be
ports
like
you
know:
icp
mp6,
multicast,
listener
discovery,
all
kinds
of
stuff
would
turn
back
on
and
would
cause
sometimes
based
on
the
timing.
It
would
cause
a
ptf
test
to
fail
that
ran
afterwards,
and
people
had
commented
on
that.
C
There
had
been
occasional,
sporadic
reports,
but
finally,
hanov
had
it
happened
to
him
as
well,
and
I
finally
decided
to
spend
the
day
and
just
dig
into
what
happened,
and
so
I
put
a
bunch
of
remedies
for
that,
so
it
shouldn't
happen
anymore
after
I
launched
hcsc
which
we're
not
even
actually
using
for
the
test.
At
this
point
I
redisable
ipv6
to
make
sure
it
doesn't
happen
again.
C
So
this
was
a
worthwhile
pursuit
and
we
finally
got
to
the
root
cause,
have
a
fix
and
then
we'll
have
an
actual
improvement
in
the
in
the
generator
itself.
So
awesome
yeah.
Sometimes
when
you
take
the
time
you
can
get
to
the
root
cause,
it's
great
because
you
come
up
with
a
really
good
fix
that
you
can
sleep
with.
C
I
would
welcome
one
if
someone
wanted
to
take
the
time,
I'm
not
in
a
rush
for
this.
If
someone
wants,
does
anyone
want
to
wait
till
next
week
to
try
this
or
should
we
just
go
for
it?
I
feel
pretty
comfortable
about
it.
C
Are
there
any
I'm
sorry,
I
didn't
hear
you.
B
Oh
I'm
saying
I'm
okay
with
it
uh-huh
chris.
I
have
just
one
question,
very
speaking,
yeah.
So
what
problem
does
this
request
result?
There's
no
easy
packets.
C
Okay,
the
the
problem-
I'm
not
sure
if
you
heard
my
earlier
explanation,
but
it
if
you
were
just
to
let's
say,
start
with
a
make
clean
and
and
then
run
all
the
steps
that
are
documented
in
the
readme,
there's
a
good
chance
that
the
first
time
you
run
make
run
all
tests.
C
C
B
So
I
understand
so
it
looks
like
this
is
not
actually
the
so
we're
trying
to
fix
the
ipv6
package,
but
probably
some
other
noises
can
happen
in
the
future
anyway
right,
but
why
we
do
not
write
the
test
cases
in
the
manner
that
will
filter,
filter
out
the
package
that
you
know
we
just
expect.
Okay,
we
can.
We
could.
C
We
could,
but
you
might
actually
conceal
a
real
problem
someday
if
you
just
start
filtering
things
right.
That's
like
oh
I'll,
delete
everything.
I
don't
want
to
see
and
look
for
the
thing
I
expect.
That's
that's
a
recipe
for
hidden
bugs
right,
so
we
really
need
to
be
scrupulous
and
try
to
make
sure
things
work
exactly
as
we
expect
and
we
we
get
exactly
the
packet
we
want
and
nothing
else
and
try
to
make
sure
we
that
if
we
have
a
contaminated
test
environment,
which
I
would
consider
this
contamination,
we
try
to
clean
the
contamination.
C
B
Okay,
yeah:
I
agree
that
we
can
disable,
but
anyway
we
once
we're
reading
the
test
cases
which
should
you
know,
verify
that
if,
if
you
send
the
packet-
and
you
expect
that
it
will
not
be
received
by
another
site,
please
verify
that
this
just
this
packet
and
that's
it.
So
in
this
case
you
can
filter
out
all
the
rest
of
the
pack.
So
you
know
we
cannot
just
rely
on
the
ips6,
probably
some
other
packets,
maybe
because
it's
linux
environment
right.
So
we
can
get
a
lot
of
packets
like
yeah.
C
C
No,
it's
not
unexpected
vxlan
packet,
it's
it's!
Linux,
ipv6
discovery,
packets!
Coming
from
linux
control,
plane,
you
know,
saying:
hey!
Look!
I
see
a
b
port,
I'm
going
to
start
sending
queries
out
and
looking
for
things
right.
It's
this
normal
linux,
behavior
and
the
typical
remedy.
What
I've
seen
in
you
know
these
kinds
of
let's
say
tutorial
projects,
etc,
etc.
C
D
Right
so
since
we
are
getting
that
packet,
how
is
it
you
know
causing
the
test
to
fail.
C
It's
causing
it
to
fail,
because
at
the
time
that
packet
comes,
which
is
random
right
then,
at
that
moment,
ptf
test
case
was
running
that
sent
a
vxlan
packet
into
the
data
plane
and
expected
the
vxlan
packet
to
come
back
modified
and
it
didn't
get
that
vxlan
packet.
It
got
a
linux
control,
pane
packet.
C
A
C
The
first
run
make
all
run
all
tests
because
that's
when
xcsc
got
enabled
and
started
allowing
linux
to
leak
packets,
so
just
kind
of
a
test
environment.
You
know
hygiene
issue
and
it
may
not
be
gone
forever.
D
So
so
then
we
just
wait
until
the
vxlan
packet
comes
back
and
it
does
come
back
at
some
point
right.
The
expected
vxlan
packet.
D
D
At
start
yeah,
as
you
said,
you
know
in
the
future,
we
might
need
ipv6,
etc.
So
it
might
be
good
to
wait
for
multiple
packets
until
the
packet
we
want
to
see
is
also
received.
Maybe
that's
something
we
can
discuss
offline.
C
D
C
Yeah,
we'll
probably
need
to
you,
know
brainstorm
a
bit
as
a
group
and
figure
out
what's
the
right
way
to
handle
this
in
a
linux
environment
right.
What's
what's
the
preferred
way
because,
like
I
said,
we
may
end
up
covering
up
a
real
problem
and
leaving
you
know
technical
debt
around,
so
it's
probably
making
a
tempest
in
a
teapot
right
now
about
this,
but
now's
the
time
to
kind
of
you
know,
discuss
things
where
we
we
don't
want
to
build
in
technical
debt.
B
Yes,
it's
an
interesting
issue
to
be
honest
because
I
tried
like
to
this
case
a
few
times
and
send
the
vxlan
and
expect
some
modified
packets
with
very
noisy
environment.
I
I
understand-
probably
I
can
add
some
reference.
Maybe
later
I
think
how
we
handled
so
maybe
it
can
help.
C
Yeah
yeah,
and
maybe
we
decide
you
know
as
a
group,
all
right-
we're
going
to
run
things
through
a
filter
because
we
know
this
is
going
to
happen,
and
but
we
want
to
do,
we
want
to
make
sure
we're
not
covering
up
real
problems
right
and
just
getting
nervous
about
patches
and
band-aids,
because
they
catch
up
to
you
later.
The
worst
time.
A
C
A
Yeah
github
is
so
picky
and
mario
did
you
want
to
talk
about
this
on
marion
or
are
you
is
this
for
another
time.
B
Oh,
that's
probably
for
another
time
I
don't
know,
maybe
we
can
review
it
tomorrow
on
the
behavioral
model.
Okay,.
A
Okay
sounds
good
good,
so
this
is
what
I
have
for
today.
Unless
we
had
some
q
a
to
happen
or
something
you
guys
need
from
me.
D
Just
want
to
update
that
we
tried
chris's
two
or
two
er,
the
the
ci
voldemort
mechanic.
Would
you
like
to
say
anything
about
that.
C
Oh,
you
know
what
there's
there's
one
item:
that's
not
showing
up
in
this
list
yet
because
I
haven't
filed
a
pr
and
that's
migrating
everything
to
to
azure
container
registry,
I'm
having
to
work
on
a
branch
to
do
that,
because
you
need
the
permissions
of
the
repo
to
be
able
to
push
to
the
docker
hub.
But
I
will
be
once
some
of
these
other
prs
merge
I'll,
reconcile
everything
with
maine
and
then
we'll
be
discussing
the
switch
over
to
docker,
I
mean
to
add
your
container
registry.
C
You
have
to
look
at
the
right
branch
to
even
see
that
work,
so
I've
been
filing
prs
to
that
branch,
which
you
mentioned
christina
the
prs
were
closed.
What
they
are
is
just
incremental
prs
to
that
branch
in
order
to
make
progress,
but
on
a
public
pr.
Yet
it's
just
like
work
in
progress
that
I
have
to
do
it
that
way
in
order
to
have
the
authentication
credentials
to
do
the
publishing
of
the
dockers.
But
let's
say
in
the
next
couple
weeks.
I
would
expect
this
whole
migration
to
be
complete
and
that'll.
Be
nice.
C
C
And
then,
like
I
mentioned
last
week,
I'll
document
the
workflow,
it's
kind
of
a
maintainer
workflow.
It's
not
you
know
day-to-day
developer
thing.
It's
a
few.
People
need
to
be
able
to
learn
how
to
do
that.
So
we
have
a
team
approach,
so
I'm
not
the
only
one
who
knows
how
to
do
it,
we'll
cross
that
bridge
when
we
come
to
it.
A
Good,
I
wonder
if
at
some
point,
we'll
have
a
split
into
a
full-on
test,
testing
work
group
right
now.
I
have
it
kind
of
wrapped
into
the
overall
dash
and
then
testing
is
in
this
call
as
well.
I
didn't
want
to
have
too
many
calls,
but
maybe
some
point
in
the
future.
We
we
have
a
like
a
testing
only
where
we
dig
into
the
guts
of
it.
C
Yeah,
well,
things
are
going
to
heat
up
pretty
soon.
You
know
we're
starting
to
get
enough
critical
mass
here
where
test
cases
are
going
to
eventually
start
rolling
in,
and
maybe
that
will
be.
The
right
call
decide
as
a
t
as
a
group
here
how
we
want
to
do
that.
Maybe
we
alternate
every
other
week
or
something.
A
A
A
So
just
the
last
thing,
then,
you
know
if
I'm
ever
to
go
out
of
office
or
get
hit
by
a
bus.
Reshma
has
kindly
offered
to
host
the
meeting
for
me.
So
thank
you.
Reshma.
A
Or
if
anyone
else
wants
to
as
well,
you
know
feel
free.
Just
let
me
know
share
it
share
it
out,
but
that's
all
I
have
for
today
and
let
me
see
if
there's
any
new
people
on
the
call.
I
don't
see
anyone
new,
so
I
will
give
you
time
back,
except
for
maybe
chris
or
marcia.
Can
you
stay
on
the
line
and
help
me
figure
out?
This
word
list
that
I
edited.