►
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
The
first
section
is
a
conversation
basically
just
to
introduce
ourselves
and
figure
out
what
we're
doing
here
together.
So
I
have
some
prompts.
A
I
maybe
go
around
the
room
and
say
your
name:
what
team
you're
on
whether
that
is
a
team
in
a
larger
organization
or
if
you're
in
a
small
organization,
the
whole
organization,
maybe
what
something
you
don't
have
to
explain
what
you
do,
but
maybe
something
short
if
you,
if
folks,
if
you
want
to,
if
you
are
working
directly
with
ipfs
in
some
fashion,
maybe
how
are
you
moving
data
around
on
ipfs
or
other
networks
today?
What
would
you?
How
would
you
like
to
be
moving
data
around?
A
What
are
your
pain
points
right
and
then?
Finally,
what
brought
you
to
this
session
like?
What
are
you
hoping
to
get
out
of
it?
And
what
what
do
you
want
to
learn?
And
you
have,
I
think,
we'll
we'll
do
like
a
minute
30
for
each
person,
I'm
not
going
to
put
a
whole
timer
up
there,
but
like
if
you're
going
ridiculously
long,
maybe
I'll
start
making
faces
at
you
cool
all
right.
Everyone
ready
to
anyone
want
to
get
started.
I
mean
I
can
just
start
with.
Are
you
okay
to
start
yeah?
B
Like
web
restoration,
nft
storage
analysis
and
I'm
just
getting
to
the
point
where
I'm
starting
to
have
close
to
plastic
with
us
and
try
to
be
useful,
I
think
the
way
that
it
moves
data
around
is
the.
B
As
implemented
on
aws
lambda
and
like
awsc
services,
I
don't
know
how
to
do
that.
I
think
graph
sync
looks
cool
honestly.
What
I'm
going
to
take
it
out
is
I'm
killing
time
before
the
like
real
privacy
tracks.
C
Hi
yeah
I'm
daniel.
I
also
joined
recently
two
months
ago
about
as
a
developer
advocate
for
ipfs
and
I
am
working
on.
C
With
bit
swap
and
http,
I
built
a
really
lightweight
implementation
of
like
bfs,
and
actually
what
I'm
really
hoping
to
get
out
of
this
is
to
understand
more
the
trade-offs
and
like
graph
sync
and
versus.
G
E
Team
right
now
I'm
moving
data
around
vlogging
on
file,
pointing
I
guess,
I'm
working
on
the
vm.
So
I'm
not
really
moving
data
around
replicating
the
data
and
kind
of
regenerating
it,
but
I
do
usually
like
book
questions.
That's
all!
Yes!.
A
E
But
yeah,
so
what
I
want
to
go
this
morning
is
to
explain
all
the
problems
with
bits
off
and
I'll
think
like
basically
where
we
want
to
take
it,
nothing
to
be
able
to
learn
and
how
we
think
it
like
worked
better
because,
like
right
now,
it's
kind
of
like
it
has
literally
been
worked
on
by
like
four
or
five
people
over
time
with,
like
they
we've
been
sort
of
changing
their
standing
possibilities
to
work
over
time.
A
H
F
Consensus
lab
at
field:
we
are
moving
data
in
a
horrible
way
because
we
just
use
dp2p
and
gossip
shop
to
like
our
class
as
part
of
player
and
yeah.
We
want
to
understand
what
extent.
H
F
Between
our
browsers
and
our
server
nodes-
and
that
is
all
of
our
scientists,
we
would
like
to
move
it
around
a
little
bit
faster.
We.
D
F
Latency
problem
there
like
very
neat
bags,
so
I'm
very
interested
in
grassing.
A
G
Hi
I'm
model
at
the
stewards
right
now.
I'm
moving
data
around
with
the
swag
hdp.
G
Like
protocol
and
also
into
the
http,
they
would
be
just
regular
protocols
and
https,
and
I
hope
to
get
out
what
I
hope
to
get
from
this
market.
This
one.
A
A
I've
moved
a
lot
of
data
around
with
volt
bitswap
and
grassic
and
worked
on
both
libraries
a
little
bit,
and
I'm
really
glad
you're
here
to
my
talk
is
also
about
how
graphsync
has
a
million
problems,
so
it's
gonna
be
like
problems
from
from
stephen
about
bitswap
and
then
problems
from
me
with
grassy,
but
what
I
hope
to
get
out
of
this,
I
would
like
folks,
here's
here's
what
I
want
I
want.
A
If
you
are
here
for
the
whole
thing,
but,
but
something
that
I've
noticed
in
working
on,
I
guess
really
only
two
of
the
protocols
is
that
there's
some
common
questions
that
come
up
and
some
common
ideas
and
I'm
gonna-
spend
like
10
minutes
before
steven's
talk
just
sort
of
like
identifying
what
I
think
I
know
which
are
like
cross-cutting
concerns
that
you
will
encounter
no
matter
what
protocol
you
are
going
to
use.
A
So
I'm
gonna
this
is
gonna,
be
like
very
short
intro
to
the
whole
track,
so
I'm
gonna
there.
These
are,
I
think,
there's
like
five
of
these
and
and
I'll
just
go
through
them.
Yeah
key
concepts
all
right,
so
this
is
concept
called
incremental
verifiability.
A
To
me,
this
is
like
the
baseline
to
call
yourself
a
protocol
labs
protocol
and
it
should-
or
it
should
be
essentially
like
we're
moving
around
content
address
data
you
should
generally,
you
should
generally
be
not
care
who
you're
getting
it
from,
because
the
content
address
data
should
be
self-verifying
and
we
should
fetch
it
in
a
way
like
we
don't
have
to
verify
every
piece
of
data
we
get
bite
for
bite.
A
We
can
hold
some
untrusted
data
for
some
amount
of
time,
but,
like
you,
shouldn't
have
to
download
100
gigabytes
before
you
figure
out
if
you've
got
the
right
thing,
and
this
would
be
like
like
if
we're
gonna
use
http
as
a
protocol,
I
would
love
it
if,
like
to
what
you
were
saying
like
you,
could
do
it
in
such
a
way
that
you're
not
just
waiting
until
the
very
end
to
figure
out
if
you
got
the
right
thing.
A
So
that's
that's
a
concept
called
incremental
verifiability
and
it's
like
you'll
notice
that
most
of
the
protocols
have
to
think
about
this,
like.
How
are
we
gonna?
Do
it,
given
what
we're
planning
to
do?
How
are
we
gonna
accomplish
this?
A
Another
concept
I
want
to
like
this
is
a
word
that
I
use
a
lot
and
I
don't
think
anyone
else
uses,
but
which
is
that
I
call
it
query
planning
which
is
you're
most
likely
trying
to
get
a
large
piece
of
data,
possibly
a
whole
dag,
possibly
a
part
of
a
gag,
and
possibly
you
want
to
get
it
from
one
or
more
people
and
maybe
over
one
or
more
protocols.
A
A
Very
few
of
our
protocols
do
this,
but
it's
a
really
important
question
and
there's
a
lot
of
solutions
like
you're
gonna
hear
about
like
some
there's
a
section
on
this
sort
of
idea
of
manifest,
which
might
be
a
really
useful
way
to
think
to
be
able
to
split
up
queries
among
multiple
peers,
but
there's
a
lot
of
things
you
a
lot
of
different
approaches.
You
might
try
like
you
might
want
to
move
like
if.
A
A
large
unix
fs
file,
it
might
make
sense
to
only
ask
one
person
via
a
graph
sync
query,
for
the
entire,
like
superstructure
of
unix
fs
file
down
to
the
blocks.
This
is
assuming
we
don't
solve
the
block
limit
problem,
but
and
then
go
to
bit
swap
for
all
of
the
raw
blocks
at
the
bottom,
which
is
where
the
large
bits
of
data
live.
A
So
that's
just
an
example
of
a
way
you
could
plan
a
query
shared
data
model,
so
this
is,
I
think,
one
of
the
most
interesting
concepts
when
you
think
about
a
transport
protocol
is
what
do
both
sides
both
have
to
understand
about
the
data
to
move
it
a
bit
swap
is,
is
probably
the
one
that
that
has
the
most
straightforward
thing,
though
possibly
sending
htv
cars
is
even
simpler
because
you
don't
even
know
their
cars
until
the
very
end
but
like
if
you're
just
sending
blocks.
A
All
you
have
to
know
about
is
that
you
have.
You
have
bytes
that
hash
to
sids
when
you
move
into
protocols
like
graphsync
or
any
of
their
or
any
protocol
that
operates
above
the
block
layer.
You're
asking
both
sides
to
understand
the
data
at
a
higher
level
and
and
if
you
came
to
the
data
model
data
models
track
yesterday,
something
that
came
up
a
lot
is
that
we
don't
all
agree
on
how
we
should
understand
data
above
the
block
level
right.
A
So
there's
an
interesting
question
here,
like
both
like
how
what
do
you
both
have
to
understand
about
the
data
and
like
how
do
you
maybe
signal
to
each
other
things
about
the
data
so
that
they
can
so
that
the
other
party
can
understand
like
there's
this
there's
a
super
advanced,
you
know,
plans
somewhere
in
the
protocol.
Labs
caves
to
like
send
wasm
blocks
around
to
help.
A
You
understand
new
ipld
data
structures,
like
that's
pretty
wild
to
me,
I
don't
know
like,
but
that
sounds
like
a
lot,
but,
like
I'm
just
saying
you
can
take
this
really
far
right
and
it's
it's
an
interesting
thing.
I
don't
know
what
the
last
word
on
that
slide
is.
I
finished
this
very
soon
recently
how
how
okay
this
I'm
just
calling
this
I
have
this
category
called
server
concerns
right,
which
is
like
if
you're
writing
a
protocol.
A
A
How
do
you
minimize
your
resource,
consumption
and
then
the
flip
side
of
that
is:
how
do
you
make
sure
data
is
moving
over
the
wire
constantly
and
not
you're,
not
just
leaving
connections
idle
and
how
are
you
saturating
connections,
and
then
is
this
the
last
one:
oh
yeah,
no,
no
and
then
the
last
one
I
have
is
like
I'm
vaguely
calling
this
error
recovery.
A
It's
not
really
a
category,
but
there
are
some
common
things
that
come
up
a
lot
which
are
like
things
that
happen
that
go
wrong
drop
connections
right
errors
sending
on
the
pipe
that
is
not
you'd,
be
surprised
how
how
annoying
it
is
to
deal
with
that,
and
that
is
a
reality
just
if
you're
writing
a
new
protocol
on,
particularly
if
it's
on
live
pdp.
A
You
should
be
for
the
time
being
aware
that
your
connection
might
drop
and
have
a
plan
to
recover
missing
or
partial
data
right
like
a
lot
of
a
lot.
We
have
these
big
dads.
A
lot
of
people
do
not
have
all
of
them.
How
are
you
dealing
with
that
there's?
There
are
some
problems
around
like
race
conditions
and
data
consistency.
A
We
have
a
little
bit
of
an
easier
time
with
this,
because
we
are
dealing
with
content
addressed
immutable
data,
so
you
kind
of
have
a
very
good
indicator
if
you're
out
of
date
on
data
but
there's
still
stuff
like
okay,
I
requested
a
block,
and
now
I
already
got
it
from
someone
else.
Do
I
want
to
try
to
cancel
it?
Do
I
want
to
you
know:
is
this
want
old?
How
do
I
not
like
get
backed
up
with
old
wants?
This
comes
up
in
bitswap,
sometimes
so,
yeah.
C
A
Broad
categories
that
I've
noticed
come
up
across
protocols
and
I
think
it's
useful
to
have
like
some
conceptual
shared
language,
to
to
help
us
talk
through
like
what
are
our
different
alternatives.
A
So
that's
all
I
got
and
that's
the
end
of
the
intro
and
now
we're
going
to
hear
from
steven
about
bitswaft,
oh
by
the
way
after
stephen
I'm
going
to
talk
about
grassland,
and
then
we
have
more
presentations.
We
have
a
presentation
about
car
sinking,
I
believe
yes,
hopefully,
hopefully
the
hopefully
yes,
okay,
cool
and
then
talking
about
a
little
bit
about
this
pattern
of
manifests
and
then
finally,
we
have
an
open
discussion,
because
that's
the
easiest
way
to
have
a
smooth
session
is
to
let
people
debate
cool.