►
From YouTube: 2018-04-03 Meeting of the DSpace Entities Working Group
B
Following
thank
you
all
for
being
here
following
our
prior
meetings
on
our
last
meeting.
We
have
showed
cap
and
that
my
proposals
to
include
entities
in
the
space
corn
and
today's
agenda
is
to
discuss
and
analyze
the
the
third
option
based
on
items
versus
the
space
object
based
based
version.
So
just
a
little
tour.
B
B
On
the
option
free,
it
is
more
simple:
it
is
quick
quickest
to
implement
it.
Reuses
lots
of
items,
solutions
like
metadata
support,
search,
discovery,
workflows,
item,
submission
forms,
versioning
handles
away,
Oh
a
I
support.
Rest
crude
operations
creates
tights
reads,
but
it
also
has
some
cons,
our
version.
It's
it
is
much
complex.
It
is
it
it
is
more
difficult
to
implement.
It
requires
more
resources
to
implement
it,
and
if
we
like
use
features
like
versioning
on
entities,
for
instance,
we
have
to
implement
it
on
on
displays
objects.
B
B
B
At
the
the
presentation
layer,
how
can
I
assign
an
entity
to
value?
Should
we
use
the
authority
control
for
that,
and
it
isn't
clear
for
me
so
I
left
some
questions
to
discuss
if
he,
if
he
choose
entities
at
item
level.
This
third
option
can
we
later
migrate
already
established
data,
for
example
using
Flyway
TV
or
something
like
that
to
to
migrate
data
that
is
already
stored
in
the
database
and
another
question:
do
we
really
need
to
generalize
this
face
object
in
which
particular
cases
do
do?
We
need
this?
B
C
Thanks
for
the
clear
overview
we
do
have
to
discuss
the
the
columns
that
you
list
sit
there
a
little
bit
in
detail
because
they're
not
all
100%
correct,
but
before
I
go
into
that
on
your
last
slide.
I
think.
Another
important
thing
that
needs
to
be
discussed
is
why
we
need
hard
coded
Java
classes
to
represent
entities,
because
both
these
base
Chris,
as
well
as
your
option
both
have
that
generic
object
in
it
an
entity
and
I
forget
the
name,
jdoah
object
or
something
like
that
at
an
Indy
space,
Chris's
solution.
B
D
B
B
C
And
I
think
at
the
end
of
the
last
meeting,
I
also
raised
the
question
of.
Can
somebody
give
me
an
example
of
a
feature
that
requires
any
entity
like
project
or
author
to
be
hard-coded
java
class
instead
of
type
of
an
item?
But
no
examples
were
given
or
of
the
use
cases
like
that,
and
we
were
thinking
and
we
went
through
a
number
of
potential
candidates
for
use
cases,
but
we
couldn't
come
up
with
any
so
that
is
really
necessary.
E
C
C
But
to
me
it
does
make
sense
that
all
objects
are
in
some
kind
of
collection,
because
in
this
phase
the
collection
is
not
just
a
way
to
group
things
together,
but
also
to
provide
some
administrative
functions
for
the
objects
that
belong
to
that
collection.
Workflow.
The
submissions
that
are
there
so
I
think
it
makes
sense,
but
it
can
be
improved
in
future
versions
or
should
be
improved.
The
Third
Point
not
a
flexible
way
to
relate
entities
by
example.
All
publications
will
have
relations
to
authors.
I,
don't
completely
understand
what
you
mean
by
them.
B
B
E
There's
there's
actually
a
minimum
amount
of
related
objects
configured
in
that
same
X
and
also
you
can,
for
instance,
states
that
a
publication
has
relations
to
others,
and
you
can
say:
okay,
there's
gonna
be
at
least
one
order,
or
you
can
also
configure
to
say
it's
gonna
be
at
least
0
Walters,
implying
that
it's
okay
to
have
a
publication
with
without
an
author.
You
can
also
say
this
at
most
one
author
or
at
most
infinite
of
authors.
C
B
C
B
That's
what
that
was
what
I
am
trying
I
tried
to
put
here:
the
net.
Okay,
you,
you
have
some
type,
some
some
type
of
workaround
solution
for
this
was
what
was
I,
trying
trying
to
say
about
this
flexibility.
If
you
have
publication
a-
and
you
want
different
relationship
from
for
different,
then
then
the
publication
B.
That
was
why
I
wrote
this
but
variability
there
are
some
workaround
solutions
for
this.
Thank
you.
C
A
Right
so
you
could
define
one
publication
type
is
like
a
book.
A
book
has
authors,
whereas,
like
a
piece
of
artwork
has
artists
and
you
could
define
each
of
those
as
separate
types,
so
there'd
be
artwork
as
a
type
and
book
as
another
type,
but
this
basic
example
was
just
for
a
generic
publication,
so
you're
right.
The
basic
example
just
shows
that
all
publications
in
this
case
would
have
authors,
but
that's
just
an
example
of
how
you
could
implement
a
generic
publication.
B
A
I
think
it's
great
that
you,
you
wrote
these
down.
Paul,
oh,
don't
don't
don't
take
it
any
other
way!
I!
Think
it's
wonderful
that
just
to
dig
into
here
and
places
that
it
might
be
confusing
or
that
that
need
clarity,
because
it's
good
to
talk
through
this
to
make
sure
we're
all
on
the
same
page
with
how
this
model
would
work
essentially.
C
C
C
But
that's
not
very
difficult
to
do.
Actually
I
mean
it's
just
extending
the
submission
forms
with
one
new
input
type
that
allows
you
to
search
and
you
would
use
solar
for
that
in
the
items
of
a
particular
type.
Yes,
overly
you
say:
okay,
we
want
to
search
for
items
of
type
author
and
a
search
can
be
done
relatively
easy.
An.
A
E
A
And
I
was
gonna
note
here,
I'm,
assuming
that
it
did
it.
If
this
was
implemented
at
the
UI
level,
it
would
not
use
authority
control
because
it's
using
the
sequel
relationship
table
correct.
That's
correct,
okay,
so
that
second
point
is
that
it
would
not
be
in
via
authority
controlled
via
via
that
table
relationship.
F
Starting
from
the
start
from
the
question
also
from
leave,
and
if
we
go
want
to
go
on
the
space
object
or
on
item
just
a
few
point,
if
we
decided
we
need
to
go
on
item
level,
I
need
to
understand
why
we
are
from
a
community
and
what
make
community
and
collection
different
than
Alta
and
why
we
don't
want
to
have
some
feature
that
we
have.
40
item
also
applied
to
the
collection,
for
instance,
when
you
create
a
collection.
F
Actually
you
are
using
a
wizard
and
I
know
that
a
lot
of
institution
we
like
to
have
an
input
format.
It
can
be
configured
to
to
describe
whether
the
collection
or
to
to
impute
alt
information
also
related
to
the
administrative
aspect
of
the
collection.
This
is
the
the
fish
stinks.
If
we
feel
that
we
only
need
item,
maybe
we
can
also
say:
okay,
we
don't
need
to
have
community
and
collection
or
justice.
Another
example
of
altar
wine,
computing
collection
are
different
than
a
metal
or
notor
profile.
F
I
think
they
are
different
and
we
could
also
be
not
immediately
aware
of
the
use
case
that
makes
this
entity
different,
but,
for
instance,
on
a
mountain
profile,
it
will
be
very
important
relation
between
the
auto
profile
and
user
account,
because
you
want
to
give
special
permission
to
the
user
that
owned
to
profile
again.
Of
course,
you
can
serve
that
in
a
lot
of
different
way
and
if
we
prove
the
acrl
or
utilization
of
this
pace.
F
You
can
decide
about
different
strategy
of
caching
depending
on
the
entity.
For
instance,
if
you
are
dealing
with
a
repository,
you
can
explain
a
huge
number
of
items
into
all
the
way
that
the
space
work,
but
you
can
expect
that
all
the
other
entity
may
be
a
huge
in
number,
but
they
will
not
change
so
much
and
I'll
reuse
it
a
lot
of
time.
So
you
want
to
keep
a
cache
to
access
all
your
auto
profile
or
your
subject
to
quickly
retrieve
this
information.
F
A
A
Maybe
they
should
just
be
a
collection,
so
there's
lots
of
different
ways
that
we
can
improve
the
data
model
going
forward
and
lots
of
things
that
I
think
we
agree
on
here,
but
I
think
when
we
talk
about
moving
everything
to
the
D
space
object
level.
That's
where
I
start
to
get
a
little
bit
concerned,
because
there's
a
lot
of
things:
D
space
object
that
are
not
item,
collection
or
community.
A
So
when
you
move
something
up
to
the
D
space
object
level,
you're
now
dealing
with
now,
you
can
create
bundles
via
input
forms,
which
sounds
a
little
weird.
You
can
create
bitstreams
via
input
forms.
Maybe
that's
a
little
bit
more
logical.
You
can
create
sites
via
input
forms.
That's
really
odd.
You
can
create
groups
via
input
forms.
That's
still
a
little
bit
odd,
because
they're,
mostly
just
names
and
relationships
and
permissions
and
Everson's,
which
is
more
an
account.
A
It
doesn't
mean
that
if
we
decide
that
things
are
at
an
item
level
now
that
they
have
to
be
at
that,
and
these
dspace
Aard
space
9,
it
could
be
that
we
refactor
this
so
that
input
forms
now
get
shared
across
items
and
collections
or
or
whatever,
and
maybe
there's
other
aspects
of
items
that
also
get
applied
to
collections,
but
to
to
argue
that
everything
must
go
to
the
d
space
object
level.
Simply
because
there's
a
similarity
between
item
and
collection
is
a
little
odd
to
me.
I'm
just
trying
to
get
my
mind
around
it.
A
E
I
completely
agree
with
that
Tim
and,
and
also
on
the
other
part
that
I'm
the
rails
suggested,
for
instance,
if
we're
talking
about
items
representing
an
altar
and
we're
talking
about
granting
permissions
for
that
author
object
to
de
person
representing
that
I.
Do
that's,
of
course,
very
easy
based
using
items
christian,
you
can
just
grant
write
permissions
on
that
specific
author.
That's
one
of
the
important
aspects
of
or
important
advantages
of
reusing.
That
item
object
you
can.
It
can
also
use
a
Smither
ID
based
available
on
the
item,
to
link
it
to
de
curse.
E
So
all
of
that
is
already
available
and
regarding
caching
I
would
say
it's
a
great
advantage
that
we
have
caching
implemented
for
items
and
how
often
does
an
item
change?
How
often
does
a
publication
change?
How
often
does
an
author
change
a
publication
once
published?
Doesn't
change
that
often
an
author
once
present
in
the
system
it
may
change
when
the
person
leaves
or
when
the
person
changes
departments.
It
doesn't
change
that
often
I
think
that's
true
for
many
objects
and
it's
very
much
dependent
on
exactly
what
your
repository
is
being
used
for.
A
F
But
typically
you
need
to
submit
a
publication
and
you
need
to
provide
a
lot
of
information
and
you
are
several
step
before
them.
The
publication
is
finalized
and,
depending
on
to
use
case,
you
can
also
face
a
scenario
where
you
have
a
thousandth
of
concurrently
submission
in
a
day,
and
if
you
try
to
make
any
cash
on
so
fastly
changing
entity,
it
will
become
a
mess.
So
you
need
to
be
able
to
distinguish
between
something
that
are
under
submission
and
change
frequently
and
something
that
has
been
finalized
again.
F
I'm,
not
saying
that
this
is
not
possible.
If
you
just
an
entity
I'm
just
arguing
that
it
could
be
more
complicated
and
the
same
recommendation.
It
is
same
that
we
say,
for
this
is
just
a
fist
version
and
we
can
start
using
a
single
entity
and
move
in
separate
entity
later
we
can
work
also
in
the
opposite
way,
so
we
can
decide
to
have
a
separate
entity
just
for
all
the
profile.
F
Now
Rafi
looks
to
be
the
most
urgent
needs
and
mesh
back
to
the
item
when
we
are
sure
that
is
just
fine
to
use
a
single
entity.
Another
thing
is
again
with
Tim,
for
instance,
about
the
importance
of
some
feature
on
other
entity.
If
you
do
to
a
person
entity
that
could
be
the
most
range
compared
to
the
item
having
out
workflow
process,
a
to
a
person
could
be
very
useful
because
it
will
mock
registration
process
so
that
you
can
decide
to
have
a
referral
to
validate
a
new
new
user
or
you
can
decide.
F
You
don't
have
these,
but
this
will
flow,
so
the
user
will
become
immediately
active
or
tingling
deaths,
having
the
ability
to
configure
to
form
the
input
form
and
everyone
workflow
is
something
that
you
want
to
apply
to
all
the
entity
that
you
manage
in
this
place.
That
utterly
is
what
we
call
displaced
object,
and
this
also
apply
to
boundary,
depending
on
which
uses
you
to
the
of
the
bundle
but
having
an
ability
to
create
bundle
to
have
overflow
for
a
pre-approval
to
edit
additional
meter.
That
bundle
level
would
be
a
great
functionality
in
some
scenarios.
A
Yeah
I
think
that
I
can
I
can
understand.
Looking
at
individual
use
cases
there
to
bring
features
to
other
dspace
objects.
I
guess
I'm
just
very
concerned
about
doing
this
sort
of
work
to
move
everything.
The
D
space
object.
If
it's
going
to
require
us
to
do
massive
changes
elsewhere,
I
feel
like
we're
almost
building
an
entirely
new
system
here.
If
we
do
that,
because
we're
gonna
have
massive
user
interface
changes
with
how
how
to
manage
bundles
the
input,
forms
and
workflow
processes,
and
now
II
people
via
that
I'm.
A
Just
worried
about
the
complexity
of
this,
so
I'm
not
saying
that
we
couldn't
eventually
get
there.
It's
possible
that
yeah,
given
enough
use
cases
and
enough
scenarios
Andreea.
You
may
be
completely
correct
that
maybe
the
maybe
a
lot
of
these
things
could
be
useful
across
other
D
space
objects.
I,
guess
I'm,
trying
to
find
a
way
that
we
can
do
this
incrementally,
rather
than
all
at
once,
and
turn
this
into
a
really
really
massive
overhaul
of
the
entire
system,
trying
to
kind
of
yeah.
F
F
So
we
need
to
be
able
to
create
a
collection
if
you
be
able
to
create
a
community,
and
so
on,
so
I
will
prefer
to
invest
one
more
day
to
have
a
general
solution
that
applied
to
all
this
waste
object
so
also
to
community
and
collection,
then
build
something
specific
for
community
collection
that
we
need
to
withdraw
in
one
release.
So
there
are
some
stuff
that
need
to
be
built
and
we
can
just
build
on
this
base
object.
Also,
because
is
not
you
that
there
is
nothing
after
this
base
object
level.
F
This
is
something
exists:
how
to
display
API
level
and
much
more
can
be
done
with
very
little
effort.
That
is
actually
what
was
done
in
this
place.
Chris
in
this
place,
Chris
you
can
browse
and
search
for
everything
and
essentially
the
requirements
until
this.
Basically
five
was
just
to
be
a
displaced
object,
so
the
change
is
very
limited
and
you
can
search
for
everything
and
it
will
be
a
great
feature
also
to
be
able
to
browse
black
community
or
collection.
A
But
I
think
that
that
proposal
in
itself,
just
even
trying
to
be
able
to
browse
by
each
dspace
object.
That
implies
a
lot
of
work
underneath
that
we're
not
planning
for
right
now.
That
implies
to
me
that
we
need
to
have
different
indexes
and
things
of
that
nature
at
the
API
layer,
not
at
the
rest
layer,
not
at
the
UI
layer
to
be
able
to
suddenly
bring
these
other
features
up,
and
that's
where
I
start
to
get
concerned.
I
feel
like
work
works,
baby.
F
Say
that
the
work
to
do
that
is
very
limited,
because
we
have
already
done
that
in
this
best
place,
and
we
need
to
do
these
similar
work
force
in
the
space
saver.
In
any
case,
because,
for
instance,
we
want
to
have
more
space
item
and
go
to
item
index
alt,
so
you
can
search
also
for
workspace
icon
of
workflow
item.
F
F
A
A
Would
learn
to
see
a
proposal
for
how
that
would
work,
because
I
am
worried
about
that.
That's
exactly
what
I'm
worried
about
is
that
I
think
that
this
is
more
work
than
we
are
anticipating,
and
so,
if
I,
if
it
is
not
as
much
work
as
I
anticipate,
it
would
be,
then
I'd
be
okay
with
it,
but
I'd
like
to
see
a
proposal
for
how
we
would
move
all
of
these
features
to
the
DS
based
object
level
in
an
easy
fashion,
because
I,
don't
I
I
can't
see
a
way
to
get
there
quickly.
Myself.
C
But
can
I
ask?
Why
would
we
need
to
do
that
right
away?
I
mean
what
what
is
the
issue
with
working
in
two
steps
and
I
mean.
If
you
talk
about
two
steps,
then
you
said
earlier
Andrea
that
you
know
we
could
also
first
hard-code
everything
and
then
move
it
to
the
item
level.
I.
Don't
really
agree
with
that,
because
that's
more
code
and
more
effort
and
more
complexity
to
then
reduce
it
I
mean
typically
in
iterations.
You
would
add
more
complexity
and
only
the
complexity
that
is
needed.
C
F
Essentially,
our
proposal
to
use
just
one
single
entity
catch-all
entity,
and
you
will
immediately
faced
with
the
most
complex
scenario
where
you
need
to
merge
something
really
dynamic.
That
is
what
we
also
have
in
this
base
crease,
but
was
a
relator
limited
to
dynamic
object
and
we
have
decided
to
have
some
fierce
citizen
object
as
entity.
F
So
if
we
want
to
go
for
iteration
state,
you
have
one
better
model
that
tried
to
resolve
all
use
case
immediately,
maybe
with
not
enough
experience
or
not
perfect
way,
I
would
prefer
to
just
say
we
want
to
solve
the
most
important
use
case,
that
is
to
auto
profile.
It
was
I
lighted
in
a
way
that
is
possible,
more
much
more
complete,
so
I
refer
to
have
a
better
orchid
integration
in
the
first
iteration.
Instead,
you
try
to
to
manage
a
generic
data
model
if
we
are
not
completely
ready
for
that.
A
A
So
we
need
a
way
to
be
able
to
to
support
those.
It
doesn't
necessarily
need
to
be
completely
out
of
the
box,
but
there
needs
to
be
some
way
to
flip
a
switch
to
allow
people
to
use
that
so
their
open
air
compliance,
without
that
some
of
our
agreements
that
we're
trying
to
work
on
with
the
open
air
group
kind
of
become
null
and
void.
So
we
have
three
entities
here,
essentially
that
we're
trying
to
to
support
at
a
bare
minimum,
and
it's
perfectly
fine.
A
A
So
that's
where
there's
more
dynamic
data
model,
I
think
becomes
much
more
interesting
because
it
would
allow
people
to
become
open
air
compliant.
If
you
want
that
compliance,
if
you
don't
want
that
compliance
for
whatever
reason
you
could
either
choose
to
ignore
it
or
just
use
author
profiles,
so
that
that's
kind
of
where
this
this
this
picture
is
coming
from
a
little
bit
and
in
terms
of
I,
think
the
dynamic
relationships
I
think
aundrea.
A
That
would
be
something
we
could
definitely
learn
more
from
you,
but
I
still
don't
have
a
good
sense
of
of
the
complication
there
and
I
still
wonder
in
my
head.
If
there's
a
way
to
to
provide
entity
specific
indexing
within
solar
or
something
to
help
manage
those
dynamic
relationships,
a
little
bit
better,
even
if
the
data
model
is
highly
dynamic,
maybe
the
solar
indexing
is
a
little
bit
more
entity
specific
in
some
way.
A
I,
don't
know
right
off
the
top
my
head,
how
that
would
work,
but
but
I
think
that's
something
we
could
definitely
learn
from
you,
I.
Don't
think
any
of
these
any
of
these
three
proposals
we're
looking
at
right
now,
I,
don't
think
we
need
to
accept
any
one
of
them,
as
is
I,
think
it's
more
of
a
matter
of
which
of
these
three
seems
the
most
promising
to
start
to
move
forward
with,
and
then
can
we
come
together
as
a
community
and
decide?
How
can
we
improve
on
this?
How
can
we
build
on
this?
A
What
are
the
points
that
the
dspace
chris
team
has
learned,
that
we
can
really
kind
of
get
around
figure
out
ways
around
some
of
the
issues
that
you've
run
into
and
encountered
in
all
your
years
of
working
on
D
space,
Chris,
so
I
think
that's
kind
of
where
it
sits
out
right
now,
it's
not
that
one.
We
have
to
basically
cut
down
each
of
these
proposals
on
their
cons,
it's
more
about.
F
So
I'm
sorry
I
cannot
give
a
order
suggestion
other
than
that,
but
I
really
it's
hard
to
me
to
to
understand
that
motor
entity
is
an
item.
It's
just
something
that
I
conceptually
I
cannot
understood
our
different
object
of
different
purpose.
Maybe
if
we
want
to
have
a
simplification,
we
have
collection,
we
have
item
and
we
have
contextual
entity.
It
can
be
everything
else.
A
So
what
is
I
guess?
What
is
the
key
difference
between
an
item
and
an
entity?
What,
if
you
just
called
an
item,
an
entity
and
then
that
that
that
item
/
entity
could
be
either
a
normal
document?
It
could
be
a
video,
it
could
be
a
author,
it
could
be
a
art
artwork,
it
could
be
whatever
sort
of
thing
you
wanted
it
to
be
like.
A
D
F
So
I
think
that
there
are
some
functional
and
non
functional
difference
between
these
entities.
It
is
to
always
visit
on
to
the
place.
If
we
go
to
try
to
find
difference,
we
can
say:
ok,
we
just
have
a
big
table
where
we
put
all
the
stuff
and
it
will
work,
because
this
is
the
way
that
database
works.
So
when
you
need
to
date
world
you
can
just
not
have
one
single
table.
That
is
key
value,
and
you
can
put
all
stuff
that
you
want
here.
A
A
Yeah
cuz,
I,
guess
I,
guess
I'm
just
trying
to
get
a
sense
of
looking
at
these
three
models.
Is
there
one
that
stands
out
as
the
most
promising
that
we
can
work
on
like?
It
seems
like
right
now
we're
working
on
the
third
sort
of
proposal
a
little
bit
and
figuring
out
ways
to
improve
that.
If
that's
the
best
use
of
our
time,
then
I
would
like
us
to
basically
say
that
that
is
the
best
use
of
our
time.
So
we
can
say
you
know
essentially
we're
working
on
this
proposal.
A
Three
we're
gonna
change
it
into
a
community
proposal
and
we're
gonna,
improve
it
and
and
analyze
this
data
model
to
try
and
find
a
way
to
to
implement
this
better.
But
but
that's
different
than
us
continuing
to
compare
these
three
models
over
and
over
and
over
again
and
I
guess
I'm
just
trying
to
get
a
sense
of.
Are
we
comfortable
getting
to
a
point
where
we
can
select
the
best
of
the
three
and
start
to
work
on
it
and
improve
it
and
turn
it
into
a
community
model
that
we
can
build
together?
B
D
B
D
H
To
be
open-air
compliant
for
others
is
only
to
have
the
chance
to
manage
other
profiles,
and
for
me,
for
example,
is
whoever
the
most
flexible
data
model
to
use
this
base.
In
whatever
context
you
want
so
I.
Think.
First
of
all,
we
have
to
write
down,
for
example,
a
document
to
make
sure
that
all
of
us
see
the
same
aim
for
these
entities
groups.
Otherwise
it
it
can
be
useless
to
discuss
about
option
one
two
or
three.
B
B
A
So
yeah
and
some
of
those
details
are,
in
the
disparities
overview
document
I'm,
linking
to
this
section.
That
says,
background
information,
why
new
entities-
and
these
were
really
key
points
that
came
out
of
steering
group
discussions.
There's
there's
several
points
in
there,
which
are
ones
that
you
mentioned
Claudio
there.
There's
the
the
need
for
author
profiles,
the
need
for
better
integration
with
orchid.
We
want
to
be
more
competitive
with,
with
things
like
be
press,
digital
Commons,
which
actually
does
really
well.
A
With
these
researcher
pages
author
profiles,
we
want
better
alignment
with
core
next-generation
repositories,
recommendations
and
we
want.
We
want
to
be
aligned
in
with
open
air
for
guidelines
and
there's
some
other
key
requirements
which
are
listed
under
there,
which
the
steering
group
actually
did
review
and
went
through
in
more
detail
in
the
last
couple
meetings
in
terms
of
trying
to
figure
out
exactly
which
things
were
optional,
which
things
were
required
and
some
of
the
things
they
listed
in
there
were
we're
actually
providing
an
extend
ability
which
you
mentioned
where
the
entities
we
want
to
add.
A
While
right
now
they're
more
Chris
specific
some.
What
Chris
oriented
they
want
to
be
able
to
provide
the
capability
for
dspace
to
be
extended
with
other
types
of
entities
for
other
types
of
use
cases,
and
that
was
something
that
came
out
of
steering.
They
also
wanted,
ideally
this
to
be
an
incremental
approach,
not
a
gigantic
step
in
d
space.
A
Just
because
it
would
keep
people
more
knowledgeable
about
what's
underneath
d
space
and
allow
them
to
bring
their
knowledge
forward
incremental
e,
instead
of
ripping
everything
out
all
at
once,
and
they
wanted
the
solution
to
allow
for
easier
upgrades
for
existing
DS
based
sites,
specifically
in
areas
where
users
may
not
necessarily
want.
Maybe
they
don't
really
care
about
their
profiles.
A
Maybe
they
don't
care
about
open
air
guidelines
because
that's
not
how
they
use
their
repository,
but
they
can
still
just
upgrade
dspace
easily
and
almost
ignore
these
entities
if
they
don't
want
to
utilize
them
at
this
point
in
time.
So
those
were
some
of
the
things
that
came
directly
from
the
steering
group
around
things.
A
D
D
That's
even
pointed
out,
for
instance,
just
put
down
onto
paper
and
open
a
discussion,
and
maybe
a
table
would
help
such
as
the
rows
could
be
represented
by
specific
requirements
such
as
is
this
entity.
Does
this
entity
need
to
be
updated?
Frequently,
yes,
no
and
so,
and
in
the
columns
have
the
three
options
and
each
cell
should
specify
yes,
this
can
be
done.
Yes,
this
can
be
done
with
this
effort.
Yes,
this
or
no.
D
This
cannot
be
done
with
this
option
and
think
so
that
this
would
clarify
a
little
bit,
because
I'm
I
have
the
feeling
that
we
are
discussing
a
lot
about
how
to
implement
and
not
enough
about
what
to
implement
that
seemed
the
field.
I
mean
Dedes.
Of
course,
this
background
information
is
very
relevant.
It's
very
important
because
it
came
out
from
from
specific
context
where
all
these
needs
were
discussed,
but
at
the
same
time
there
are
so
high
level.
That
is,
it
is
difficult
to
define
the
best
implementation
based
on
these
I.
G
B
B
B
G
B
G
A
A
We
did
ended
with
look
at
authors.
Yes,
so
we
did
dig
into
authors
a
little
bit,
but
we
didn't
dig
into
necessarily
other
entities
and
part
of
the
reason
behind.
That,
though,
is
that
what
I
keep
hearing
from
from
steering
and
others
is
that
there
is
not
a
strong
agreement
from
all
the
various
use
cases
in
the
ways
that
these
space
use
is
used
worldwide.
There's
not
a
strong
agreement
on
the
very
specific
entities
that
everybody
wants
other
than
like
authors.
A
That's
not
the
goal
of
the
overall
D
space
system,
like
Andrea
I,
think
I
mentioned
earlier
in
this
meeting.
Our
goal
is
really
to
manage
and
deal
with
with
documents
as
that
primary
sort
of
entity,
electronic
documents
or
whatever
you
want
to
call
them
and
then
allowing
the
ability,
then
to
be
more
flexible
and
storing
other
sorts
of
things
in
your
D
space.
But
we
need
to
need
other
extensions
to
be
able
to
manage
those
even
better.
A
How
can
we
actually
implement
this
as
an
as
a
as
a
as
a
community
would
help
us
start
to
clarify
a
lot
of
these
questions
and
it
would
force
us
to
really
say:
okay,
you
know
if
we're
doing
this
in
this
way,
how
can
we
now
manage
these
dynamic
entities
a
little
bit
or
cash?
These
dynamic
entity
is
a
little
bit
better.
Can
we
do
that
in
solar
a
little
bit?
A
Unfortunately,
here
and
we're
a
little
over
time,
but
but
that's
where
I'd
really
would
like
to
see
us
go,
is
we
just
need
to
start
to
move
on
something
I'm,
starting
to
feel
that
most
people
are
starting
to
go
out
towards
the
option?
Three
that
I've
heard
but
I
know
obviously
that's
not
ideal
for
the
D
space
Kristine,
but
I'd
like
to
find
a
way
to
to
bring
this
all
together,
so
that
we
can.
We
can
make
sure
that
the
needs
of
the
dues-based
Chris
team
can
be
met
in
this
same
sort
of
situation.
A
F
So
there
is
another
big
question
in
our
document
related
to
closely
Copland
relation.
It
is
very
important,
so
I
think
that
we
need
whatever
decision
we
will
take
when
we
are
ready
to
discuss
about
that.
We
will
need
to
discuss
how
to
manage
relation,
because
I'm,
not
sure
or
better
I'm
sure
that
we
can
not
only
have
structured
relation.
It
is
what
is
proposed
being
what
option
to
an
option
tree.
When
you
talk
about
publication
and
older,
there
is
a
lot
of
complexity
hidden
into
the
day.
F
For
instance,
you
have
a
lot
of
external
altar,
and
maybe
you
don't
want
to
have
author
profile
also
for
external
people.
If
also
you
decide
to
have
minimum
author
profile,
for
example,
people
with
just
a
name.
It
is
an
important
decision.
We
can
take
it,
but
is
not
the
only
issue,
for
instance,
publication
and
auto
auto
could
be
not
only
person
that
would
be
person
and
group.
You
can
have
a
researcher
group
design
a
publication
and
if
you
just
need
to
create
a
link
to
another
entity,
you
are
not
able
to
do
that.
F
A
D
F
C
F
To
manage
external
auto,
we
need
to
manage
relation
with
organization
because
they
are
already
managing
the
space,
so
it
displays
items
you
have
the
authority
framework
that
allows
you
to
manage
these
lost
kind
of
relation
so
before
to
introduce
anything.
I
think
that
we
need
to
clarify
when
you
want
to
use
auto
to
the
framework
and
when
we
go
to
use
the
relation.
F
C
Just
two
comments:
I
mean
first
words,
not
because
the
relation
is
possible
that
you
have
to
use
it.
You
can
also
use
other
means
to
add
an
author
to
an
item
or
an
entity
and
second
about
research
groups.
I
mean
you
can
also
define
an
entity.
That's
a
research
group
and
say
that
that
research
group
has
a
relation
to
persons
or
authors
and
have
that
research
group
have
a
relation
to
the
publication.
I
mean.
C
F
Want
to
stress
this
point:
I
cannot
do
that
as
well
sure
I
know
what
we
are
going
to.
We
want
to
try
to
achieve
because
I'm
editor
of
most
of
the
documentation
that
we
do
I
have
cited
so
I'm
involved
in
next
generation.
Repository
hang
P,
editor
of
the
guideline
for
open
air
and
Chris
involved
in
all
this
stuff,
so
I
know
them
very
well.
F
It's
not
the
same
to
say
you
can
create
a
relation
with
author
and
a
creation
relation
with
group,
and
you
can
decide
which
you
want
to
use,
because
the
relation
is
just
one
and
sometimes
you
need
to
use
Auto.
Sometimes
you
need
to
use
group
and
you
need
to
mix
these
two
entities
together
and
it
is
important
you
know
if
the
author
has
been
cited
before
the
group
or
vice
versa.
F
F
I
cannot
say
that
solution
three
is
underway,
but
I
think
that
has
it
is
now
is
not
enough
could
to
be
a
step
toward
unmerge
of
the
space
crease
in
the
space
of
thing
like
that,
if
some
minimal
requirements
are
not
respected,
we
will
end
with
a
solution
where
the
space
crease
user
cannot
use
the
space
7
as
a
basis
to
build
something
and
will
need
to
say
on
something
else.
This
me
may
mean
that
we
need
to
fork.
F
We
need
to
work
on
the
space
crease
instead
of
work
on
the
space
and
we
are
going
to
subtract
resource
to
to
the
community.
So
my
goal
is
whatever
we
want
to
do.
I
want
to
do
it
in
a
direction
that
will
make
a
space
crease
on
the
space
closer,
but
don't
mean
that
this
Bayes
Network
all
the
feature
of
the
space
crease
because
they
may
be
not
apply,
but
if
I
need
to
store
information
that
is
needed
to
this
base,
crease
I
need
to
have
the
right
room
in
the
space.
F
F
You
ours,
sorry
I
am
going
to
conclude,
ok
to
a
solution.
That
is
a
justify
step,
because
we
need
to
also
provide
something
fastly
and
is
not
really
the
solution
to
long
term,
because
we
say
at
start:
we
don't
want
to
have
something
that
can
be
market
against
the
price
and
have
all
the
functionality.
That
is
the
space
crease
that
already
exists
now
and
you
can
use
now.
We
want
to
build
something
better.
So
if
we
need
to
build
something,
but
we
need
to
do
that
in
the
right
way
from
the
south.
A
And
aundrea,
we
definitely
do
want
your
expertise
in
all
of
this,
and
we
do
have
a
common
vision
here
that
I
want
to
make
sure
that
that
whatever
we're
building
is
in
line
with
what
dspace
Chris
can
use
and
systems
like
dspace
Chris,
so
that
that's
why
I
was
clearly
or
I
was
trying
to
clearly
state
that
these
three
proposals
are
not
static
in
any
way,
shape
or
form
and
no
matter
which
one
we
choose.
It's
going
to
be
a
community
project.
It's
going
to
be
input
from
the
entire
community.
A
It's
going
to
involve
a
lot
of
it's
gonna
need
a
lot
of
feedback
from
dc-based
Kristen
from
you
on
Draya,
in
terms
of
making
sure
we're
aligning
this
first
step.
Well
with
these
based
Chris,
but
I
think
it's
gonna
have
to
be
a
first
step.
We
just
need
to
make
sure
that
first
step
is
in
the
right
direction
and
not
going
in
a
way
that
is,
that
does
not
align
with
dspace
Chris,
but
the
reason
why
I
think
it
has
to
be
a
first
step
is
because
of
the
the
timelines
on
these
things.
A
We
don't
want
DS
p7
to
take
forever
and
and
I'm
hearing
that
we
would
really
like
this
to
be
a
part
of
D
space
7,
if
possible,
at
least
from
steering
that's
what
they
would
really
like
to
see
and
that's
what
the
open
air
project
really
wants
to
see.
They
don't
want
us
implementing
open
air
for
and
2020.
They
want
it
by
2019.
A
If
we
can
do
that,
so
so,
I
think
it's
gonna
have
to
be
a
first
step,
but
we
want
to
make
sure
that
first
step
is
in
the
right
direction
and
in
a
way
that
we
can
build
upon
that
to
move
us
in
the
direction
of
DS
based
Chris,
so
that
the
two
roadmaps
come
together.
So
I
think
we
agree
on
a
lot
here.
Andrea
I
think
it's
a
matter
of
trying
to
figure
out
how
we
can
modify
one
of
these
three
proposals
to
better
align
with
the
needs
of
D
space,
Chris
and
similar
systems.
A
So
yeah,
that's
that's
kind
of
I
hope
you
understand
that
and
agree
with
that.
Andreea,
but
I.
Definitely
don't
want
dspace
Chris
to
go
a
different
direction.
I
agree
completely
that
we
need
to
bring
this
stuff
all
together
and
make
the
right
first
step.
I
just
want
to
make
it
clear
that
none
of
these
proposals
are
static
and
none
of
them
will
be
accepted,
as
is
they
will
all
require
community
feedback
and
ongoing
work.
F
F
A
That
right,
yep,
yeah
I,
think
steering
is
steering
agrees
with
that
completely
I.
Don't
think
steering
wants
dspace
Chris
to
be
a
separate
product.
They
want
it
to
move
closer
to
being
more
of
an
easy
add-on
to
dspace
some
people
who
wanna
can
almost
flip
a
switch
and
turn
it
on
so
I.
Think
all
of
us
have
that
same
idea
in
mind.
It's
just
a
matter
of
getting
there
and
finding
a
way
to
align
these
roadmaps.
I
think.
F
That
we
should
be
also
clear
with
the
steering
and
the
community
about
their
expectation,
because
we
spent
five
meeting
to
analyze
existing
code
and
seeing
them
of
an
existent
system,
which
is
this
best
place,
to
try
to
understand
a
bit
better.
What
is
best
wishes,
and
also
after
this
demo
there
is
not
a
wide
understanding
of
what
is
discretion
is
and
how
it
is
implemented.
F
So
the
reality
is
that
some
time
someone
will
devolve
till
today
for
today
to
develop
something
and
if
we
are
in
the
final
rush
of
aversion,
I
really
want
to
avoid
that
a
last-minute
contribution
will
be
included
because
just
we
have
promised
used
to
stealing
to
the
community
to
have
something
like
that.
More
or
less
is
what
we
say
in
turns,
light
look
reasonable
and
we
got
yourself
because
we
just
need
to
summarize
the
space
Christine
turns
light.
I
can
do
that.
F
A
A
We
can
define
what
they
are
in
the
roadmap,
but
we
can
write
them
into
the
roadmap
and
say,
and
in
dspace,
eight
we're
gonna
make
this
next
step.
That
will
help
us
align
even
better
with
these
space
Chris,
but
I
think
that
first
step
is
a
very
is
very
important
to
scope
small
because,
as
you
noted,
there's
a
lot
of
complexity
around
all
of
these
topics.
So
we
need
to
keep
the
coats
cope
quite
small,
and
we
need
to
make
it
in
the
right
direction
so
that
we
can
build
further
steps.
On
top
of
that.
A
That's
kind
of
where
I'm
coming
from
here
I,
don't
think
we're
going
to
be
able
to
solve
every
single
problem
in
one
year,
but
I
think
we
need
to
make
a
one
good
step
in
the
right
direction
and
build
upon
that
build
out
a
roadmap
for
what
comes
after
that
and
I
hope
that
you
would
agree
with
that
and
that
you
can
help
us
figure
out
what
that
first
right.
Small
step
is
and
how
we
can
start
to
do
that
alignment,
realizing
that
we
can't
do
everything
at
once.
F
F
A
So
we're
we're
well
over
time
here,
but
I.
Don't
know
that
we've
come
to
a
exact
conclusion,
but
I
think
I
still
feel
like
what
I'm
hearing
is.
People
are
leaning
more
towards
the
option,
3
proposal,
but
obviously
there's
a
lot
that
needs
to
be
discussed
in
moving
that
forward
and
making
that
more
of
a
community
project,
and
the
steering
group
has
a
meeting
tomorrow.
A
Actually
that's
their
next
meeting,
so
they'll
be
talking
about
this
a
little
bit
more
tomorrow
as
well
know
that
wall
so
I'm
sure
wrap
over
into
the
Duras
based
summit,
which
is
next
week.
So
they'll
probably
be
discussion
at
the
Duras
based
summit
as
well
and
I.
Don't
think
I
have
anything
else
to
wrap
up
with
here
today.
Unfortunately,
I
can
report
back.
A
Excuse
me,
I
can
report
back
on
what
goes
on
in
the
steering
group
tomorrow
and
also
give
a
give
a
report
back
from
the
summit,
although
I
won't
be
there
in
person,
but
I
plan
to
attend
virtually.
So.
Thank
you
all
for
your
time
and
for
the
discussion
and,
let's
figure
out
a
way
to
move
this
together,
to
get
move
this
forward
together.
Thank
you.
Thank.