►
From YouTube: CRDT Research Meetup // CRDTs in Peer - Pedro Teixeira
Description
Originally recorded during the Lisbon Hack Week from May 21-25, 2018.
A
Hi
everyone
Pedro
I
work
for
protocol
labs,
I'm
the
captain
of
this
kind
of
new
group,
which
is
called
dynamic
data
and
capabilities
working
group,
so
we're
on
github
and
our
our
main
purpose
is
to
research.
Do
research
and
development
of
the
building
blocks
that
will
enable
the
creation
of
the
centralized
apps
so
researching
things
like
stuff
for
dynamic
data
like
charities
or
identity
access
control,
all
the
primitives
that
you
need
to
build
decentralized
applications
on
top
of
IP
FS.
A
A
Currently,
there's
a
bunch
of
current
endeavors
are
more
engineering
focused,
so
JavaScript
libraries
for
charities
pierced
our
platform
peer
pad,
will
I
think
we'll
what
recipe
pads
later.
One
also
mentioned
peer
pads.
So
these
are
all
all
charity
related,
there's
a
bunch
of
open
source
that
that
we
have
published
so
connectors
for
well-known
charity.
Libraries
in
JavaScript,
so,
and
also
mainly
the
work
that
we
are
doing
is
JavaScript
bias
because
we
want
to
enable
decentralized
web
applications
to
be
developed.
A
So
you'll
see
a
lot
of
JavaScript
here
so
stuff
that
guarantees
encryption
at
rest
for
operation
by
charities,
peers
charities,
the
library
for
causal
oppas
charities,
they're
based
on
immutable
data,
so
it
kind
of
fits
well
in
the
ipfs
model,
so
transport
for
for
that,
through
the
a
BFS
network
in
implementation
of
Delta's
charities.
All
of
these
are
listed
on
the
dynamic
data
and
capabilities
working
group
websites.
A
If
you
want
to
check
it
out,
also
one
has
mentioned
peer
pads,
so
this
is
like
one
of
the
one
of
the
products
that
well
we
want
to
create
more
of
these
products
to
test
all
the
things
that
we
are
building
underneath.
Pier
pad
is,
is
the
first
and
and
as
an
engineer,
we
have
a
bunch
of
challenges
that
that
I
would
like
to
share
with
you,
I
like
to
share,
as
Evan
mentioned
most
well.
I
have
a
bunch
of
them
in
the
working
group
issues
repo.
A
A
There
are
three
dilemma
domains
as
I
mentioned
for
this
group,
so
it's
dynamic
data,
identity
and
access
control.
Right
now,
we're
still
more
focused
on
dynamic
data,
so
creating
strong
eventual
consistency,
primitives
for
real-time
collaborative
web
apps
that
are
in
loosely
connected
peer-to-peer
networks,
so
to
make
a
portfolio
of
charities
to
make
them
available
to
developers.
A
Here
the
main
challenges
are
several,
so
it
could
be
fastboot,
garbage
collection,
persistence,
replication
performance,
network
efficiency,
the
persistence,
replication
performance,
a
network
efficiency,
we're
solving
it.
That
BFS
layer
and
all
the
rest
is
it's
kind
of
dependent
on
researching
on
the
researcher,
so
I
mean
the
engineer:
are
not
a
researcher
I'm
just
putting
this
these
problems
out
to
the
research
community,
so
I'm
going
to
describe
the
two
RFPs,
the
problems
behind
it
to
RFPs
that
we
have
published.
A
A
A
A
That's
one
problem!
The
other
problem
is
local
garbage
collection.
So
how
does
the
replica
safely
compact?
The
the
history
is?
This
two
are
kind
of
related
to
problems
that
are
kind
of
related,
so
we've
bundled
that
up
into
an
OPA
problem
statement,
that's
optimized,
storage
and
convergence
time
in
causal
commodities.
So
I
encourage
you
to
take
a
look
at
it
at
that.
A
So
and
so
there's
an
Associated
RFP
as
Evan
described
the
other
one
is
more,
is
related
to
dynamic
dilemmas
but
more
about
access
control
of
eventual
virtually
consistent
data.
So
how
do
we
create
a
cryptographic,
ACL
type
that
Garrett's
still
guarantee
strong
eventual
consistency
when
applied
to
charity?
A
So
I
think
this
is
a
a
heart
problem,
because
if
you
think
that
you
can
have
like
offline
nodes
that
are
performing
changes,
they
think
they
have
the
capability.
But
now
the
capability
has
been
revoked
by
by
some
other
node
that
has
that
capability
of
revoking
and
still
when
I
come
up
when
I'm
connected
to
the
network.
A
Again
I
still
have
a
bunch
of
operations
that
I
want
to
send
to
the
other
nodes
either
the
other
all
the
other
nodes
know
that
I
don't
have
the
capability,
and
so
that
that's
that's,
okay,
no,
no,
no
other
node
will
accept
those
operations
or
if
this
is
not,
they
are
not
consistent
about
about
that
knowledge.
They
will.
Some
of
them
will
take
the
operation
and
the
others
may
not
accept
that
operation,
so
they
have
different
views
on
the
ACL
at
a
certain
time.
So
this
is.
A
This
is
something
that
that
I
think
we
I'd
like
to
have
more
insight
on
from
the
research
community.
So
there's
an
again
there's
a
no
problem
statement
describing
this
and
there's
an
Associated
RFP
that
describes
this.
Then
we
have
a
bunch
of
other
problems
that
I
could
I
could
go
on,
but
I
won't
I,
don't
have
the
time.
So
if
you
want
to
chat
about
this
letter
or
head
out
to
these
arab
repos
and
the
research
repos
repo
in
protocol,
that
evan
shows
has
this
RF
peace.