►
From YouTube: OpenActive W3C Community Call / 2020-11-18
Description
Accessibility Specification
- Proposal
- Modelling
A
Okay,
welcome
to
the
call
for
the
18th
of
november
2020,
we'll
just
be
talking
about
aspects
of
the
accessibility
specification,
or
rather
the
bit
of
the
opportunity
specification.
That's
concerned
with
accessibility,
I'll
just
start
showing
the
screen
here.
A
C
A
Has
that
improved?
Can
you
see
the
slides
now
yep
yeah,
okay,
great
okay?
So
these
are
not
really
about
the
the
meat
of
the
accessibility
specification.
This
is
more
kind
of
one
might
say:
sort
of
admin,
details
or
standards,
details
about
the
standard
and
formal
characteristics.
I
suppose.
A
There's
three
components
to
it,
one
of
which
is
simply
transport.
Note,
simply
because
research
indicates
that
information
about
in
particular,
public
transport
is
quite
valuable
for
people
with
accessibility
needs.
Accessibility,
support
level
is
a
small,
controlled
vocabulary
with.
I
think
three
values
indicating
how
much
support
people
can
expect
if
they
attend
the
opportunity
and
then
most
of
the
meat
of
it
is
in
the
accessibility
support
object,
which
is
an
array
of
complex
objects,
each
of
which
the
intention
is
to
describes
some
particular
accessibility
feature
in
a
fairly
high
degree
of
detail.
A
So
the
idea
is
that,
if
you
had,
I
don't
know
four
or
five
accessibility
features.
You'd
have
four
or
five
of
those
objects
in
that
array.
So
the
open
questions
with
regard
to
this,
the
first
one
is
just
about
information
repetition.
If
you
look
at
the
example,
you
can
see
that
there
is
a
slot
there
in
accessibility,
support
for
contact
point
and
again
this
was
something
that
kind
of
was
was
flagged
up
by
sport.
A
England,
research,
particularly
that
often
people
with
accessibility
needs
really
want
to
talk
to
somebody,
or
at
least
email
somebody
and
get
a
very
detailed
sense
of
what's
involved
and
get
some
reassurance
that
they'll
be
they'll,
be
their
needs
will
be
met.
A
Similarly
carer
policy
for
people
who
need
somebody
to
be
in
attendance
at
the
time,
it's
important
to
to
know
whether
that
person
needs
to
pay
and
so
on
and
so
forth,
and
indeed
it's
sometimes
a
venues
that
are
interested
in
supporting
people
with
accessibility
needs
will
often
have
you
know,
free
amenities
for
for
carers,
and
that
kind
of
thing,
so
it
would
even
be
an
advertising
point.
A
The
difficulty,
I
guess
is
that
this
could
be
attached
at
a
couple
of
points
in
the
hierarchy
and
yeah.
It's
kind
of
it
could
get
extremely
repetitive.
So,
for
instance,
I
was
looking
at
the
active
places.
The
way
they
represent,
disability
information
or
other
accessibility
features
and
they've
got
a
fairly
fine-drained
data
that
they
capture
you
know,
it'll
be.
Are
there
accessible
toilets?
Is
the
parking
accessible?
Are
all
the
doorways?
You
know
level
nine
particular
features
that
they
model
for
every
leisure
center,
which
would
be
eight
or
nine
accessibility.
A
Support
objects
in
that
array,
but
presumably
in
many
cases
there'd
only
be
one
contact
point
for
all
of
those
or
there'd
be
at
most
two
and
that's
the
policy
would
be
global
across
all
of
those.
So
I'm
just
not
too
sure
how
to
slice
the
pie
there.
A
Yeah,
I
guess
that's
the
question.
I
guess
that
raises
two
issues,
and
one
would
be
if,
in
the
case
where
there's
one
contact
point
across
all
of
them,
that
works
fine.
What
if
you've
got
two?
I
suppose
that
if
you,
if
you
had
say
somebody
who's
in
charge
of
sensory
disabilities
and
the
support
there
and
mobility
disabilities
and
support
they're
lining
them
up
to
be
a
bit
difficult.
C
Well,
I
I
yeah
it's
in
it's
yeah,
I
suppose
there's
an
interesting
thing
here
about
like
the
level
of
kind
of
like
how
much
we
care
about
making
this
data.
Like
you
know
super.
C
I
guess
I
guess
it's
like
like
overfitting
the
model
almost
like
is
there
risk
of
that,
for
example,
is
it
really
the
case
that,
like
in
most
cases,
you
just
want
to
kind
of
list
of
the
accessibility
contacts
anyway,
because
there's
usually
going
to
be
one
or
two
of
them
and
they
probably
both
do
each
other's
jobs
a
little
bit
anyway,
and
maybe
they
job
share,
and
you
know
it
from
a
practical
perspective.
C
If
there's
a
team,
they've
probably
got
a
contact
point,
you
know,
there's
so
or,
like
you
know,
single
a
single
number.
You
can
call,
or
maybe
a
shared
email
address.
So
maybe
you
know
if
yeah,
maybe
maybe
we
don't
need
to
worry
so
much
about
that.
Maybe
you
know,
there's
just
an
array
of
contact
point
and
you
just
make
sure
that
instrument
available
or
contact
typers
you've
got
there
that
you
just
say
what
what
that
is,
and
then
they
can
figure
it
out
themselves.
C
A
C
But
that's
it
like,
I
suppose
the
question
here
really
is
what
you
know:
what
the
vet,
what
do
we
by
adding
the
extra
granularity
here
to
the
fields?
What
are
we?
What
are
we
enabling
to
happen?
So
I
guess
the
driving
force
between,
but
behind
a
lot
of
the
fields
that
have
been
added
over
time
to
the
kind
of
beta
and
things
like
that
is
usually
being
not
around.
We
need
more
free
text,
but
actually
we
need
there's
a
specific
type
of
searchable
thing,
or
you
know
something
that
has
has
meaning.
C
That's
that
that
needs
to
be
added.
You
know
like
a
video
or
something
we
can't.
We
can't
deal
with
within
the
current
thing,
but
any
extra
free
text
is
just
kind
of
yeah.
Just
stick
that
in
description.
Stick
that
in
the
additional
information
fields
so
yeah.
I
guess
to
that
end-
maybe
maybe
something
similar
applies
here.
It's
like
if
we're
adding
additional
fields.
C
Maybe
it's
just
thinking
about
each
one
really
kind
of
conservatively
about
whether
that's
that's
like
what
it's,
if
it's
necessary,
what
the
value
it
adds
over.
You
know
what
we're
going
to
write
in
the
description
anyway.
Sorry,
that's
a
slight
tangential
point,
but
that's
yeah
so
and
therefore
things
like
contact
point
yeah.
We
don't
need
to
be
so
specific
about.
A
A
C
Based
on
that
right,
exactly
that's
the
thing!
So
yeah,
it's
it's
it's
almost!
Yes,
there
are
yeah.
I
mean
if
it's
just
two
freight
free
text
fields,
for
example,
then
the
people
who
are
presenting
the
information
need
to
think
about
how
to
separate
those
out
and
provide
headings,
and
things
like
that.
Yeah.
C
So
it
could,
because
free
text
is
just
a
bit
more
awkward
to
to
present
really
if
you've
got
one
description
field
that
can
be
formatted,
that's
quite
nice
kind
of
open
to
most
people's
whatever
they're
capturing.
But
anyway
that's
yeah.
That's
an
aside!
So
but
there's.
B
A
Yeah,
the
kinds
of
the
kinds
of
remediations
you
make
for
somebody
with
say
social
or
behavioral
needs
are
so
different
from
what
you
make.
Yes,
mobility
needs
it
it
does.
It
is
very
blobby
that
way,
yeah
and
in
a
way,
it's
kind
of
like
it's
almost
like
html
divs
is
what
you're,
what
you're
describing
in
the
end.
C
Yeah,
and
also
of
course,
we've
got
to
be,
we've
got
to
bear
in
mind
that
there's
going
to
be
people
with
with
this
accessibility
issues,
reading
the
content
so
actually
yeah
having
it
separated,
not
formatted.
It's
probably
is
it's
probably
right:
are
they
all
the
fields
in
there?
The
name
description
are
they?
Is
that
complete,
or
is
that
more?
A
So
that's
great!
No,
it's
a
little
small
can.
A
But
yeah
it's
it's
not
it's
sort
of
intentionally
fairly
small,
because
we
actually
don't
have
a
lot
of
data
to
leverage
a
lot
of
the
time
and
we
need
to
give
people
realistic
goals
about
what
they're,
what
they're
going
to
capture?
That's
great
sure,
okay,
I
know
we've
gone
a
bit
off.
Oh
sorry,
sorry
one
one
that
isn't
on
there
is
ours
available.
A
C
A
I
think
in
that
structure
it
would
probably
be
well.
I
guess
that's
a
question
about
contact
point
then,
if
it's,
if
because
it
applies
across
several,
it
should
go
at
a
higher
level.
A
C
Oh,
no,
so
sorry,
sorry,
I
would
suggest
yeah.
So
I
was
suggesting
if
we,
if
we
went
with
kind
of
accessibility
contact
point
as
an
array
of
contact
points
at
the
like
event
level,
I
guess
and
the
same
level
as
accessibility
support
exists.
C
But
then
I
mean
I
mean
if
we
did
that
or
if
we
didn't
you've
still
got.
The
point
of
the
question
of
someone
wants
to
write
a
description
about
general
accessibility
approach
or
policy,
but
I
guess
this
is
an
array
and
it's
got
a
name.
So
if,
if
there's
a
thing
that
doesn't
really
have
a
name,
it's
just
a
general
description.
Are
you
thinking
that
it's
just
got
no
name
field,
just
a
description
single
element
in
the
array
type
thing.
A
A
Okay,
I
think
that
would
work
out
you
know
and
then
the
description
this
would
be
for
for
further
information.
You
know
we.
We
are
happy
to
support
individuals
depending
on
their
needs.
Please
contact
us,
you
know
to
discuss
this
further.
I
guess
something
along
those
lines.
Sure.
C
My
yeah
yeah
nish
has
a
power
plant
power
outage
today
in
his
village,
which
obviously
means
everyone
working
from
home
hasn't
it.
So
I
just
got
a
phone
call
from
him
and
then
it's
very
hard
to
discern
from
his
zero
communication
today,
whether
that
was
an
urgent
problem
or
what
oh
right,
it's
all.
Fine.
Sorry
can
you
please
continue.
A
Okay,
so
I
think
I
think
that's
so
your
recommendation,
I
guess,
would
be
yes
a
distinct
kind
of
contact.
Point
accessibility,
contact
point
living
one
level
up
yeah.
I
think
that
seems
that
seems
fair
and
then
I
guess
only
only
at
that
level
so
that
you're
not
parsing
it
further
down
the
down
the
tree.
C
Yeah,
I
think
so
well.
C
Alternative
is
that
you
only
have
it
at
the
next
level
down,
and
then
I
guess
I
just
don't.
I
don't
know
what
the
data
looks
like,
because
how
many
of
these
are
there
going
to
be?
I
think
I
think
that
level
above
is
probably
I
mean,
given
the
scarcity
of
the
data.
I
think
the
level
of
up
is
probably
the
right
thing.
We
can
always
add
an
extra
level
of
detail
later.
A
Yeah,
I
think
I
think
the
problem
is
actually
that
the
data
is
very
spread
that
or
rather
not
spread
but
polarized
in
the
sense
that
there's
most,
you
know,
leisure
centers,
supporting
organizations
that
don't
have
a
clear
way
to
represent
this
and
don't
really
capture
the
data
that
clearly,
or
at
least
not
in
a
standardized
way,
and
then
you've
got
institutions
that
are
specifically
dedicated
for
people
with
accessibility
needs
and
often
just
one
kind
of
need,
and
of
course
they
have
dreams
and
their
own
their
own
way
of
looking
at
it.
A
C
You
know
you
if
I
don't
know
what
type
the
accessibility
support
is
there,
but
you
might
be
able
to
just
inherit
it
from
something
that
has
contact
point
available
anyway.
You
know
so
that
it
would
be
very
easy
to
add
that
later.
If
someone
during
implementation
has
a
burning
need,
yeah
on
a
specific
level.
A
Yeah,
I
think
that
I
think
that
makes
sense,
but
keep
things
as
simple
as
possible
for
now.
A
A
So
modeling,
okay,
so
this
is
one
I'd.
Imagine
that
you've
been
over
with
lee
in
the
in
the
deep
past.
At
some
point,
this
must
have
raised
its
head.
So
there's
a
distinction
to
be
observed
between
accessibility,
information
which
is
really
about
the
event
you
know.
So,
if
you've
got
say
specific,
I
don't
know
emotional
or
physical
support
for
some
particular
class,
but
then
it's
also
in
a
building
that
is
wheelchair
accessible.
A
What
do
you
do
with
that?
Do
you
just
put
everything
into
do
you
treat
the
accessibility
as
a
feature
of
the
event?
Generally
speaking,
do
you
put
the
stuff
that
relates
to
the
location
in
the
location,
object
and
the
stuff
that
relates
to
the
event
on
the
event,
object
or
or
what
I
feel
like
for
parsing
simplicity,
you
kind
of
want
everything
to
be
in
one
place,
but
then
that
does
a
certain
violence
as
opposed
to
the
reality
of
the
situation.
C
Well,
yeah,
I
guess
they're
saying
different
things
right
so
that,
what's
on
the
event
and
what's
on
the
place,
sorry
is
that
maybe
I
missed
what
you
would.
A
So
most
of
the
most
of
the
accessibility-
that's
out
there,
accessibility,
information,
that's
out
there
right
now
is
about
places
it's
about
locations,
and
it's
it's
saying
you
know:
we've
got
step.
Free
access,
we've
got
disability,
disabled
parking.
That
kind
of
thing.
Then
there
is
a
small
amount
of
it
which
is
really
about.
If
you
come
to
this
class,
there
will
be
limited
light
stimulation
or
limited
noise
stimulation.
That
kind
of
thing,
which
is
only
about
that,
which
will
only
be
the
case
for
that
particular
hour,
right.
D
A
Now
both
of
those
can
be
fit
into
this
accessibility
support
array
right,
you
could
have
gym
quiet
time
for
the
event
and
you
could
also
say
step
free
access
or
whatever,
and
all
of
this
is
living
on
the
event
directly,
so
to
speak.
A
On
the
other
hand,
a
case
could
be
made
that
the
accessibility
support
is
really
in
the
latter
case
about
step.
Free
access
say,
is
really
about
the
venue
itself,
and
maybe
what
you
want
to
do
is
just
have
a
link
back
to
the
venue
for
that
kind
of
thing,
because
it's
going
to
be
repeated,
you
know,
for
everything
that
happens
in
that
venue
that
step
free
access
is
going
to
be
there.
So
how
do
people.
C
Retrieve
yeah,
I
think
the
reason
that
it
wasn't
it's
not
an
event.
It's
sorry,
the
reason
it's
not
in
the
reason
it's
not
in
place,
I
think,
is
because
the
type
of
accessibility
that
places
have,
which
are
things
like
ramps,
you
know,
like
physical
physical
attributes,
had
a
different
model
to
the
type
of
accessibility
notes
that
you
might
add
to
an
event,
and
I
think
we
only
focused
on
one
of
those
two.
But
given
that
we're
revisiting
this
properly
now,
it
would
seem
to
make
sense
to
do
both
because
they
are
different.
C
I
mean
whether
you,
however
you're
modeling,
that
ramp
it
needs
to
live
in
the
place
right
because
it
is
an
attribute
of
the
place
and,
however
you're
modeling,
the
you
know
the
the
particular
session
that
is
suitable
for
wheelchairs,
for
example,
wheelchair
basketball
session.
That
probably
needs
to
be
in
the
session.
A
Yeah,
I
guess
the
question
is
thinking
about
it
from
the
point
of
view
of
like
an
end
user
or
or
an
intermediary
like
parasport,
I
mean,
I
guess
what
you
want.
Is
you
want
to
do
a
search
on
everywhere.
That's
got
stuff
free
access,
so
you've
got
a
you've,
got
a
wheelchair-bound
user
who's
like
okay.
You
know
what
can
I?
What
can
I
use?
Yes,.
A
C
Yeah,
I
think,
yeah
exactly
if
you're
looking
for,
but
but
then
this
leads
on
to
a
slightly
separate
question.
So
in
the
in
the
mcr
project.
At
the
moment,
we've
got
things
like
pool
hoists,
which
are
obviously
an
attribute
of
the
swimming
pool
and
the
way
that
they're
modeled
using
the
spec,
which
conforms
to
the
2.0
spec,
is
by
extending
the.
What
is
the
name
of
that
thing?
C
Oh
amenity,
a
yes
amenity
feature
exactly
that.
So
we've
we've
got
additional
amenity
features
defined,
which
are
like
the
pool
hoist
because
because
this
spec
allows
you
to
extend
the
minity
feature
to
add
custom,
custom
ones
and
that's
super
useful
because
they
want
to
do
kind
of
filtering
by
these
things.
And
that's
that's
the
use
case
around
places
is
more
filtering
wise.
C
What's
going
on,
so
I
kind
of
wonder
whether
well
I
don't
I
don't
know,
but
I
wonder
whether
there's
because
obviously
I
haven't
I'm
not,
as
I
said
at
the
beginning,
I'm
not
the
expert
on
this
probably
need
to
make
sure
that
we've
got
parasport
in
this
conversation,
but
for
sure
in
like
use
cases
in
terms
of
use
cases
but
yeah.
I
think
I
think,
because
they're
different
use
cases
it
may
be
worth
thinking
about
that
from
a
modeling
perspective
as
well
like.
Is
it
the
same
model?
B
C
C
You
could
put
a
lot
of
things
in
there.
We
just
haven't
formalized
it,
but
I
don't
know
it
all
comes
back
to
these
cases.
A
So
if
you
did
do
a
amenity
feature
that
way,
I
think
you'd
want
a
very
fine-grained
kind
of
controlled
vocabulary
to
describe
those
things
which
shouldn't
be
hard
to
create
in
principle
that
you
know,
there's
there's
some
finite
number.
I
mean
it's
more
than
most
people
would
imagine,
but
you
know
it's
it's
under
10
say
for
each
of
those
and.
C
Well,
look,
I
guess,
is
the
question
a
use
case.
One
does
does
that?
Does
the
user
want
to
search
by
these
things
and
if
they
do,
it
needs
to
be
a
controlled
vocabulary
somewhere,
because
that's
how
they
can
unless
it's
a
keyword
search,
but
that's
probably
not
going
to
you-
know,
allow
them
to
display
the
information
and
the
icons
on
whatever
it
is
they
want
to
display?
A
A
A
Yeah
yeah
and
again
I
guess
that's
the
point
of
the
contact
point.
I
guess
is
if
you
once
you've
narrowed
it
down,
you
can
get
that
information
by
other
means.
C
Yeah
yeah
I
mean
it
would
it
would
be
really
good
if
we
do
have
another
call
on
this
topic.
Just
I
know
parasport
specifically
very
interested
in
this
stuff.
You
know
having
having
someone
from
them
on
this,
because
I
know
it's
quite
technical
about
schemers,
but
it's
very
difficult
to
always
to
talk
about.
A
Okay,
so
then
we
we'd
actually
be
looking
at
using
amenity
feature
for
places
and
then
more
dynamic
kinds
of
support
would
be
going
into
into
the
event.
Then.
A
C
I
guess
you
could
do
both
for
the
accessibility
stuff.
I
mean
if
it's
helpful
again
in
mcr,
they've
got
a
place.
Sorry,
a
page
for
a
place
and
they've
also
got
a
page
for
an
event.
There's
a
separate
pages,
so
you're
looking
at
the
place,
you're
looking
at
all
the
things
that
are
in
the
place,
you've
got
the
description
and
the
image
all
the
details,
the
video
in
some
cases
about
that
space,
which
is
like
a
leisure
center
or
whatever,
and
then
that's
where
the
pool
voiced
information
sits.
C
You
know-
and
this
has
got
these
these
this
place-
has
you
know
something
called
gin,
blah
blah
blah
and
it
has
poi
hoist
so
so
yeah
I
mean
you
can
imagine
accessibility,
information
sitting
there
as
well
as
you
could
on
the
event
page.
A
C
Yeah
and
like
I,
but
I
think
that
they'd
be
well
up
for
engaging
in
this
level
of
discussion.
We've
had
them
on
this
call
before
and
we've
had
other
we've
had
the
last
access
political.
I
think
we
had
accessibility,
alliance,
power,
support
activity,
yeah
activity
alliance,
yeah
exactly
and
then
and
a
couple
other
organizations
someone
from
sporting,
but
maybe
so
yeah.
I
think
I
think
parasport
were
there,
but
anyway,
yeah
they're
definitely
they've
been
in
this
forum
and
I'm
definitely
able
to
contribute
in
this
forum.
For
sure.
C
So
I
mean
yeah
definitely
worth
because
we
don't
have
to
have
we
can
we
can
gauge,
we
can
make
the
technical
conversation
as
technical
as
it
needs
to
be.
I
mean
sometimes
the
way
those
conversations
have
worked
is
that
we've
we've
kind
of
had
we've
bounced
between
the
use
cases
and
the
tech
in
the
same
call,
so
we
kind
of
do
a
little
bit
of
use
case
and
then
okay,
but
this
could
look
like
this
in
the
schema.
Oh,
but
does
this
then
work?
C
If
this
happens,
you
know,
and
we
can
kind
of
cross
check
the
scenarios
with
them
as
we
go
and
get
something
that's
kind
of
got.
I
guess
broadband,
I
suppose
I'm
a
bit
yeah
kind
of
thinking.
If
we
make
any
decisions
off
the
back
of
this
call,
it
sounds
like
maybe,
if
they've,
if
they've
narrowed
the
use
cases,
then
maybe
that's
something.
A
No,
that
makes
that
make
sense,
okay,
so
the
next.
The
next
point
was
I've:
just
not
done
an
action
to
talk
to
them
about
that.
So
this
is
about
data
providence.
A
A
A
C
C
I
yeah,
I
don't
know
if
I'm
trying
to
think
there's
any
other
aspects
of
the
data
that
really
that
kind
of
latency
applies
to,
but
I
guess
not,
because
most
things
are
very
current,
or
at
least
most
things
are
able
to
be
updated
by
anyone
in
the
marketing
department
who's
trying
to
keep
the
data
current,
whereas
this
particular
information
might
be
a
bit
more
yeah
yeah
yeah,
there's
already
a
field
in
the
booking
spec
that
we
use
for
the
terms
and
conditions
to
see
when
they're
updated,
which
I
imagine
is
the
same
field
from
schema
that
you're
looking
at
so.
A
Okay,
maybe
just
a
line
there.
I
guess:
what's
the
guidance
on
that,
I
mean
to
really
be
robust,
it
would
have
to
be
required
and
in
principle,
there's.
No.
That
should
be
one
of
the
easiest
pieces
of
data
to
supply.
A
Yeah
the
providence
information
I
mean,
I
worry,
that
data
data
entry
people
are
going
to
see
this
as
the
least
important
and
least
interesting
thing
to
enter
it's
going
to
be
an
afterthought
for
them,
but
in
fact
it's
well.
It
makes
the
rest
of
it.
C
You
don't
have
to
go
out
and
collect
it
right.
That's
true!
I
I
was
thinking
it
was
a
system,
level
property,
but
yeah.
I
see
what
you're
saying
could
be
entered
manually
as
well,
because
if
you've
got
a
system,
that's
capturing
this
data,
then
obviously
the
last
time
that
field
was
updated.
A
Yeah,
but
I
think,
if
you
imagine,
a
sort
of
active
places,
style
workflow
right
where,
where
there
is
a
kind
of
manual
check,
yes
or
indeed,
social
prescribing,
I
think
that
kind
of
manual
check
is
going
to
be
very,
very
common
as
well.
C
Yeah,
fair
enough
also
the
yeah
well,
the
property
I
was
looking
at
is
called
date
modified
from
the
from
schema,
so
yeah
I
mean
without
again
knowing
much
more
about
the
domain.
It
sounds
like
a
useful
property
to
include
sure
I
guess
they
don't
need
to
include
it
if
they
don't
have
accurate
data.
C
C
We've
definitely
had
people
put
dummy,
and
example,
values
in
you
know
their
data
that
doesn't
just
to
pass
the
validator,
which
is
completely
not
a
point
so
but
then
I
mean,
I
think
this
is
where
your
you
know.
The
profiles
that
are
being
developed
come
in
potentially
right.
So
if
you've
got
some
accessibility
profile,
then
maybe
this
is
a
score
on
that.
C
So
it's
not
required,
but
profiles
is
a
bit
of
a
softer
approach
to
that.
Rather.
C
Yeah,
because
I
think
also
what
we
seem
to
have
learned
with
required
fields
is
it's.
The
required
fields
are
more
kind
of
system
level.
Things
like
what
is
required
to
make
this
make
sense,
because
if
you,
if
you
start
straying
outside
of
what
is
required
from
a
system
perspective
like
you,
need
the
activity
list,
property
for
example,
that
has
to
be
there
because
from
a
system
perspective,
you
can't
show
it
on
the
search
results,
if
it's
not
there,
so
that
has
to
be
there.
C
But
if
you
start
mandating
fields
that
you
know
are
not
that
like
image,
for
example,
then
you
can't
you
can't
really
because
it's
basically
the
difference
between
data
validity
and
data
quality
and
the
required
fields
should
probably
be
around
the
validity
of
the
thing
I
mean.
Is
this
piece
of
data
useful
enough
to
be
used
anywhere?
And
if
it's,
if
it's
bad
quality,
that's
almost
separate
again,
I
guess.
Maybe
the
profiles
cover
that.
A
D
Right
so
yeah,
okay,
yeah!
That's
a
that's
a
good
point.
A
And
then,
finally
modularity,
so
this
is
a
pretty.
This
arose
out
of
a
discussion
with
somebody
working
for
a
comaball,
which
is
an
organization
who
worked
with
airbnb
for
getting
accessibility,
information
for
airbnb
properties,
and
he
said
in
his
view
that
actually
all
sectors
had
roughly
the
same
accessibility
problems
and
that
it
was
sort
of
slightly
artificial
to
segment
out
one
particular
sector.
So
I
suppose
that
raises
the
question.
A
I'm
kind
of
doing
this
back
to
front
here,
but
should
this
be
something
that
gets
taken
to
schema.org
as
a
kind
of
general
policy?
They
don't
have
anything
right
now
covering
this
domain.
C
Yeah
yeah
yeah
yeah,
well,
yeah.
I
mean,
I
think,
if
you
ping
the
mailing
list,
you'd
get
a
really
yeah.
You
get
them
interested
in
this,
if
indeed,
yeah
and-
and
I
think,
if
maybe
the
thing
to
do
with
that
is-
propose
something
like
a
little
a
little
concrete
and
and
then
I
mean
maybe,
and
just
this
is
just
a
suggestion
that
they
seem
to
the
way
the
schema
philosophy
seems
to
be
draw
on
the
best
from
lots
of
different
different
areas
and
then
bring
it
all
together.
C
So
they
don't,
they
don't
specialize
anything
themselves
they're
about
what's
the
best
in
so
I
guess,
if
there's
no
standard
for
this
at
all
at
the
moment
that
we
can
like
lean
on
and
say
we're
drawing
from
this
existing
thing,
then
maybe
the
best
we
can
do
is
you
know,
get
get
all
the
parties
we
have
together,
get
them
to
agree
on
something
that
looks
pretty
credible
and
put
that
into
a
github
issue,
or
you
know,
maybe
maybe
a
markdown
file
in
the
repo
or
something
and
then
post
that
into
their
mailing
list
and
say
look.
C
This
is
a
complete
x.
We
were
interested
in
in
whether
would
you
be
interested
in
bringing
this
kind
of
work
into
schema.org
is
there?
Is
there
benefit
in
that?
Because
the
other
thing
is
that
obviously
schemas
developing
stuff
that
describes
real
things,
but
they're
also
now
describing
events
and
things
that
aren't
sorry
describing
things
that
aren't
digital
anymore.
A
I
guess
the
the
difficulty
from
a
sort
of
program
perspective.
I
guess,
is
that
this
is
going
to
be
a
long
process.
I
can
imagine
this
will
go
through
iterations
and
a
lot
of
changes
before
you
know
it
would
ever
become
part
of
schema.org.
C
Well,
I
think
I
think
yeah
and
but
I
think
that's
why
you
know
us
prioritizing
getting
our
duck
scenario
first
and
presenting
up
the
finished
product
to
them,
because
they
they
never.
If
you
kind
of
watch
the
mailing
list
discussions,
they
never
engage.
People
in.
Oh,
have
you
thought
about
this
because
of
your
date.
You
know
your
domain
blah
blah
right,
they're,
not
saying
you
know,
I
think
they
should
be
pool
hoists
or
not
like.
C
They
have
no
idea
they're
just
talking
about
it
from
a
kind
of
modeling
perspective,
and
how
does
it
fit
in
with
other
areas
that
they've
already
modelled?
And
you
know,
do
you
want
to
prefix
that
property
with
something
it
doesn't
clash
with
something
else
or
whatever?
So
I
mean
sometimes
I
get
into
that
detail
but
like
it
doesn't
seem,
especially
this
with
this
topic
that
I
feel
like
we
have
to
solve
the
problem.
C
First
almost
and
then
present
something
credible
that
we've
obviously
all
agreed
on,
which
means
we
probably
would
put
that
stuff
in
beta
right
ourselves
and
adopt
it
in
parallel
and
start
to
push
it
and
and
see.
You
know
if
we
can
get
some
adoption
on
on
the
properties
as
they
are
and
then
and
then
you
know
see
if
we
can
then
do
that
in
parallel,
and
we
can
move
the
properties
out
of
beta
into
schema,
as
we
have
with
time
zone
stuff.
C
You
know
as
as
and
when
and
that
can
be
that
could
take
years,
but
but
at
least
but
that's
it,
but
it's
a
good
process
to
engage
in
and
if
we've
already
done
the
hard
work
and
we're
proving
it
every
day
that
we're
waiting
for
them.
That's
probably
the
thing:
isn't
it
rather
than
yeah.
If
we
can,
if
we
have
a
situation
where
we're
blocked
on
schema,
we're,
probably
never
gonna,
it
doesn't
work
that
way.
Does
it
schema
responds
to
demand
yeah?
So
we
can't,
as
we've
seen
with
the
api
stuff.
A
Yeah,
okay,
so
I
mean,
I
think,
that's
not
a
not
an
immediate
priority
but
yeah
once
once
road
tested
and
tires
kicked
and
so
on,
yeah,
okay,
so
having
having
dealt
with
the
big
chunky,
one
first,
starting
at
the
top.
So
right
now
this
is
part
of
the
opportunities
back.
This
is
really
yeah,
just
a
bunch
of
I
guess
it
would
be
be
hip
properties.
Should
it
actually
be
its
own
kind
of
specification,
which
I
guess
is
not
entirely
independent
of
the
schema.org
question.
A
But
you
know:
should
this
be
accessibility,
0.1
or
something
right
now.
C
C
Obviously
this
is
this
does
have
a
lot
of
overlap,
but
a
lot
of
the
things
that
already
in
the
existing
modeling
spec
we've
got
our
ducks
in
a
row
around
how
to
deal
with
multiple
versions
of
specs
now,
because
all
the
tooling
just
has
its
own
version,
so
we
can
deal
with
routes,
one
booking
one,
whatever
two
and
then
as
long
as
yeah
and
and
the
tools
just
support
the
latest
so
until
that
breaks
which
it
hasn't.
C
Yet
so
I
mean
it's,
not
it's
not
going
to
be
an
issue
logistically
having
it
in
a
separate
document.
I
don't
know
if
it
does,
it
does
get
a
bit
like
there's
lots
of
documents
to
look
at
then,
if
you're
trying
to
implement
this
and
you're
looking
for
the
kind
of
source
of
truth,
but
I
guess-
but
I
guess
what
I
was
saying
at
the
beginning
is
to
well
sorry.
C
I
was
saying
just
just
now
is
to
kind
of
getting
it
into
a
markdown
form
or
into
a
you
know
something
that's
more
kind
of
rough.
If
you've
seen
facility
types
facility,
types.
C
Yeah
so
open
active.
I
o
facility
types,
it's
very
basic
compared
to
this,
but
it's
basically
a
basically
a
repository
that
holds
a
controlled
vocabulary
and
a
beta
property
which
is
presently
adopted
by
like
a
handful
of
the
publishers,
well
actually
exactly
by
gladstone
and
the
classical
system
and
then
all
of
their
customers
but
yeah,
and
this
had
some
work
that
went
into
it.
C
There's
a
github
issue
that
relates
to
it
you
can
see
which
is
in
the
in
which
we
transferred
into
the
repository
from
the
main
repository
from
from
the
modeling
spec
repository.
C
So
so
it
kind
of
is
its
own
repo
for
the
purposes
of
one
particular
more
more
detailed
thing
which
felt
more
than
the
modeling
spec,
but
not
quite
enough
for
a
full-blown
spec.
C
So
I
mean
you
know:
one
one
route
is
to
start
with
something
like
this,
which
is,
I
guess,
a
bit
bit
briefer
and
then
then
use
this
to
surface,
to
get
feedback
and-
and
if
we
put
it
in
beta,
you
know
get
get
better
adoption.
I
think
the
problem
we've
got
is,
I
don't
think
because
of
the
nature
of
this,
I'm
not
sure
we
could
put
it
straight
into
the
main
namespace
yeah
yeah,
because
it
probably
needs.
Probably
it's
one
of
those
things.
C
Yeah
so
so
pro
I
mean
so
something
like
this,
which
would
which
is
basically
a
proposal
which
you
can
imagine
that
example,
there
being
slightly
bigger
and
covering
what
you
had
on
the
previous
slide,
and
I
don't
know
if
you've
seen
that
you
can.
If
you
go
to
open
active,
I
o
slash,
ns,
hyphen,
beta
or
even
slash
test
type
and
interface.
C
These
are
all
there's
a
we've
got
a
tool
that
you
can
use
to
generate
the
documentation
with
the
classes,
the
properties,
everything
yeah
and
the
stuff
all
comes
from.
The
json
ld
document.
C
You
could
you
know
you,
could
you
could
put
it
in
beta,
you
put
it
in
its
own
name,
space
yeah,
I'm
just
thinking
of
different
ways
of
doing
it
or
you
just
have
one.
If
you,
you
could
just
have
a
markdown
document
and
just
use
beta,
because
there's
not
that
many
properties
and
then
just
use
the
markdown
document
to
describe
those
properties
and
give
examples
of
of
their
use.
A
C
And
then,
hopefully
as
well,
that
will
mean-
because
I
imagine,
with
these
type
of
things,
that
maybe
the
implementation
of
this
will
take
a
few
months
for
people
to
finally
engage
with
it
and
what
we've
seen
happen.
Obviously,
with
most
of
these
things,
is
that
as
soon
as
the
people
that
are
going
to
do
the
work
actually
engaging
it
properly,
which
happens
you
know
with
a
bit
of
a
lag.
Suddenly
you
get
all
this
like
feedback
all
over
the
place.
Oh
well,
we
thought
about
this.
This
is
going
to
work
in
our
interface
yeah.
A
That
was
one
thing
that
came
out
really,
interestingly,
in
the
workshop
this
morning
regarding
social
prescribing
that
one
question
that
I
had
not
been
anticipating
was:
how
does
this
relate
to
snomed
the
the
medical
controlled
vocabulary,
and
I
was
like
wow,
it
could
wow
and
to
a
certain
extent,
if
you
look
at
some
higher
level
social
prescribing
sites
that
don't
ensure
that
level
of
granularity,
you
can
see
that
there
is
kind
of
a
bridge
between
physical
activity
and
this
very
granular.
You
know
very
formal
medical
taxonomy,
but
yeah
it's
a
it's.
A
You
know
it's
like
a
radical
new
area
of
work
that
opened
up
that
I
hadn't
anticipated.
So
yeah
flexibility
is
going
to
be
key
once
once
engagement
is
higher.
You
know.
B
A
I
think
the
feed
question
so
one
yeah
one
question
I
noted
down.
There
was
just
you
know:
where
does
this
live
and
but
I
guess
that's
actually
answered
a
bit
by
the
location
event.
Kind
of
dichotomy.
C
Yeah,
this
does
seem
like
it
attaches
to
existing
stuff.
Doesn't
it
rather
than
being
its
own
world
of
thing?
I
guess
a
bit
like
a
bit
different
to
roots
with
it,
which
is
like
an
entire
sphere
of
its
own,
which
is
almost
completely
unrelated.
This
is
very
much
descriptive
properties
about
the
things.
A
Okay,
I
think
I
think
that
that
one
it's
nice
to
end
on
one
of
the
simpler
questions.
I
think
that
was
where
I
was
kind
of
leaning
but
yeah.
Okay,
thank
you
nick.
That
was
extremely
useful.
I
think
you're
right
that.
Obviously
we
do
have
to
broaden
the
conversation
out
to
other
organizations
with
more
sort
of
domain
and
more
concrete
use
cases,
but
I
think,
as
general
orientation
for
the
spec.
That
was
really
helpful.
C
Oh
no
worries
look
forward
to
seeing
as
this
progresses.
This
has
been
one
of
the
things
that
open
active
has
been
trying
to
do
since
the
beginning,
so
really
great
to
see
that
the
progress
being
made.
It's
so
encouraging.
D
It's
marrying
up
the
two
sides:
yeah
yeah
yeah
yeah,
sure
absolutely
great.
Okay,
thanks
very
much
nick.