►
Description
----------------------------------------
Missed last year's IPFS þhing? Register to attend IPFS þhing 2023 in Brussels, Belgium here: https://2023.ipfs-thing.io/
A
A
So
after
that
we
thought
about
a
new
protocol.
That's
really
simple:
reframe
is
a
request
response
protocol,
so
server
and
client
working
and
talking
together
is
extensible.
So
you
can
add
more.
You
can
like
change
or
add
more
into
specifications
to
use
different
transports
and
serialization
protocols.
Right
now
we
are
covering
only
http
and
json
dax,
a
serialization
and
http
for
transport,
but
more
to
come,
and
you
are
welcome
to
add
your
own
specific
specs
into
our
specs
repository.
A
So
right
now
the
main
use
case,
as
I
said,
is
the
content
discovery,
but
it's
an
open
protocol.
So
you
can
use
this
for
any
other
use
cases,
as
you
can
use
the
frame
to
point
to
a
bitter
and
network
addresses
you
can
provide
some
refinable
http
downloads
or
you
can
create
a
private
network
with
content
providers
using
this
this
this
protocol.
A
So
right
now
we
have
covered
on
into
specs
these
methods.
They
identify
it's
a
way
to
say
from
the
server
to
say
to
the
client.
Okay.
I
support
this
this
and
that
so
after
that,
this
is
like
the
main
method
that,
to
put
all
in
the
same
page,
the
client,
the
server.
Then
we
have
the
find
providers
that
is
able
to
find
provided
theirs
by
a
cid,
and
then
we
have
the
get
ipns
to
get
some
ipns
records
and
they
put
ipns
to
put
records
into
the
into
the
server.
A
So
we
have
a
specification,
so
you
can
go
there.
You
can
read
them.
You
can
see
what's
wrong
what
you
can
improve,
do
any
pull
requests
anything
that
you
can
implement
this.
Also,
if
you
want-
and
we
have
several
frame
implementations,
the
main
the
first
one
is
a
go
delegated
routing.
That
is
basically
a
library
that
is
used
in
in
our
clients
and
servers.
A
So
this
is
implementing,
as
I
said,
the
client
and
the
server
part
of
the
reframe
protocol
store
the
index.
It's
a
as
we
say
before
in
previous
slides.
A
It's
a
server
of
the
reframe
and
protocol,
so
any
client
can
connect
directly
to
it
and
get
information
from
all
the
index
that
is
like
storing
hydra
booster
is
a
client
for
the
stored
index,
so
hybridize
the
this
dht
node.
That
is
able
to
improve
the
dsd
performance
on
the
ipfs
network
and
it's
connecting
to
the
sturdy
index
servers
using
reframe,
some
guys,
an
experiment
and
basically
it's
a
way
to
teleguide
routing
to
other
services
using
the
refrain
api.
So,
basically
right
now
it's
implementing
dht
and
store
the
index
under
the
hood.
A
So
you
can
talk
with
the
server
using
the
reframe
protocol
and
then
it's
querying
all
the
all
these
routing
systems.
Under
the
hood
and
kubo
we
did
in
the
last
in
the
in
the
last
release
candidate.
We
added
support
from
for
reframe
routing
and
you
can
try
it
now.
If
you
want
using
this
configuration
to
connect
to
the
stored
index
servers-
and
here
we
have
some
demos.
A
A
And
everything
is
working,
so
you
can
test
it
right
now,
if
you
want
going
to
the
kubo
repository-
and
here
is
another
another
example
using
some
guy,
that
is,
this
refrain
server,
but
instead
of
using
the
dht
and
the
store
the
index
under
the
hood,
I
just
hardcoded
some
peers
ids
to
see
how
simple
it
is
and
how
easy
you
can
like.