►
From YouTube: Slingshot Phase 2.3: Closing Ceremony
Description
Slingshot 2.3 closing ceremony, held on June 9th 2021.
00:00 Welcome & Agenda - Deep Kapur
1:36 Phase 2.3 Closing
4:19 Phase 2.3 Judging & Results
13:22 Phase 2.4 Kickoff
14:58 FileDrive and Go-Graphsplit - Laura
23:35 Common Crawler data processing with FilSwan - Charles Cao
35:11 Storing the A2D2 dataset - Fei Yan
42:51 Twin Quasar - Julien Noel
49:01 Q&A and Closing - Deep
A
Pretty
exciting
lineup
of
sessions
for
you
today,
so
I'm
excited
to
get
started
thanks
for
those
of
you
that
waited
and
for
those
of
you
that
are
attending
from
various
parts
of
the
world
at
a
time
zone.
That
may
not
be
the
most
convenient,
but
of
course
this
is
also
being
recorded
and
will
be
available
for
several
weeks
months
and
years
for
viewing.
But
yeah
really
excited
to
be
here
thanks
for
those
that
can
make
it
out.
A
If
you
have
any
questions
or
if
you
have
any
comments,
please
feel
free
to
drop
them
in
the
chat,
I'll,
be
monitoring,
and
so
will
the
wonderful
team
that's
here
today
with
that,
let's
keep
things
off
so
on
the
agenda
today.
We've
got
brief
sort
of
closing
comments
from
me
on
phase
2.3
as
well
2.3,
as
well
as
phase
2.4,
which
is
actually
already
underway
and
crossed
a
milestone.
Today,
then,
we'll
have
a
brief
presentation
from
laura
on
file
drive
charles
will
be
presenting
about
his
work
with
common
crawl.
A
Faye
is
here
to
present
as
well
with
his
work
with
the
a2d2
dataset,
and
then
we've
got
a
video
recording
from
julian,
since
it's
an
inconvenient
time
in
france
for
him
to
present,
but
he'll
be
talking
about
his
work
with
the
free
rainbow
tables.
Dataset,
we'll
also
try
to
keep
some
time
at
the
end
for
a
live
q
a
so.
A
You
can
also
ask
questions
of
the
participating
teams
that
are
presenting,
and
so,
if
you
have
any
questions
for
them
as
they
go
to
their
presentation,
please
feel
free
to
drop
that
in
chat
great,
let's
get
started
so
first
up.
Congratulations
on
completing
slingshot
phase
2.3!
This
is
a
pretty
exciting
phase.
We
actually
had
lots
of
data.
You
know
onboarded
close
to
nine
pebby
bytes
through
the
course
of
the
space.
You
can
take
a
look
at
this
graph
here.
A
It's
really
illustrative
in
terms
of
all
the
way
from
september,
when
sync
shot
2
started
the
way
in
which
the
rate
has
continued
to
increase
the
amount
of
data
that's
being
onboarded
onto
the
network
has
grown
substantively
phase.
2.1
was
you
know,
0.8
peppy
bytes
phase
2.2
was
over
three
per
bytes
phase.
Three
was
close
to
nine
per
bytes,
and
so
it's
it's
growing
substantively.
A
A
The
full
list
can
be
found
at
their
repo
for
slingshot
it's
somewhere
in,
like
the
mid
60s.
I
believe
at
the
moment
so
landsat
8
was
disallowed
from
slingshot
phase
2.3,
because
lots
of
teams
onboarded
it
already
and
during
the
course
of
2.3,
we
achieved
similar
amounts
of
like
lots
of
replicas
of
common
crawl,
genome
arc,
linux,
iso
and
so
moving
forward.
A
Those
will
no
longer
be
eligible,
but
lots
of
copies
of
those
available
thanks
to
the
progress
in
phase
2.3,
and
so
hopefully
we
can
find
ways
for
the
community
to
engage
with
that
data
as
usual,
we're
also
continuing
to
receive
lots
of
contributions
from
the
community.
So
if
you
have
a
data
set
that
you
think
should
be
eligible
for
slingshot,
please
feel
free
to
open
up
an
issue
or
open
up
a
pr
at
that
repo.
A
Research
science
tend
to
typically
be
the
ones
that
get
accepted,
but
you
know
happy
to
entertain
any
ideas
that
you
have
and
so
feel
free
to
create
an
issue.
If
you'd
like
to
kick
off
a
discussion
to
include
your
data
set
or
data
set,
that
you
found
that
you'd
like
to
encourage
being
onboarded
onto
the
network
cool.
So
as
we
discussed,
you
know
we
crossed
13
baby
bytes
at
the
end
of
this
phase
that
then
unlocked
a
total
rewards
pool
of
a
hundred
and
sixty
five
thousand
file
coin,
which
is
absolutely
massive.
A
25
teams
were
eligible
at
the
end
of
this.
This
includes
you
know,
teams
that
complied
with
all
of
the
rules
of
how
the
data
needed
to
be
stored,
which
meant
no
team
was
violating
the
maximum
number
of
replicas
for
each
piece
that
could
be
stored
onto
the
network
teams
were
using
at
least
four
miners
to
store
the
data
with
and
no
more
than
30
percent
of
the
data
was
being
stored
with
a
single
miner,
and
so
there
was
some
sort
of
distribution
of
data
across
miners.
This
is
a
quick
screenshot.
A
I
know
it
isn't
particularly
big
or
particularly
useful,
but
I
just
want
to
show
you
that
sort
of
what
it
looks
like.
So,
if
you
head
over
to
this
link,
it's
shared
in
the
slingshot
announcement
channel
and
if
you
haven't
seen
it
and
want
to
see
it
and
can't
find
it
feel
free
to
dm
me
on
slack,
I'm
deep
kapoor
on
slack
and
I'll.
Send
you
a
link
to
this,
but
just
sort
of
give
you
an
idea
of
how
it
it
looks
and
how
you
should
read
it.
A
So
you
have
the
project
names,
which
are
the
names
of
the
participating
teams
that
are
onboarding
data
onto
the
network.
The
do
metadata
score,
which
is
of
the
information
that
was
submitted
by
these
teams,
about
the
deals
that
they
made
the
documentation
score,
which
is
the
quality
of
the
docs
that
were
written
and
provided
by
the
team,
so
that
anybody
can
understand
how
to
use
what
they've
onboarded
onto
the
network
to
ensure
that
it's
useful
for
people
in
the
future.
A
The
deal
ui
score
so
teams
for
this
phase
actually
created
websites
for
each
of
their
projects
that
you
can
check
out
demo
videos,
so
these
you'll
actually
see
one
of
these
later
during
this
presentation.
But
in
general
you
know.
Each
team
now
has
like
a
few
minute
video
that
talks
about
the
brief
explanation
on
the
data,
their
logic
for
how
they
onboarded
content
onto
the
network.
How
you
can
find
that
and
how
you
can
retrieve
it
and
use
it,
and
all
of
the
videos
are
actually
available
on
on
youtube.
A
And
so,
if
you're
looking
for
any
links
at
the
easiest
way
to
head
over
to
youtube
to
find
the
playlist
and
if
you're
not
able
to
access
youtube,
then
there's
links
available
for
food
fee,
and
you
should
also
be
able
to
check
that
out
from
the
slingshot
website
and
yeah.
So
you
know
happy
to
discuss
this
a
little
bit
further.
If
you
have
any
questions,
I'm
going
to
dig
into
a
couple
of
the
details
here
for
those
of
you,
especially
that
are
interested
in
participating
in
future
rounds,
definitely
pay
attention
to
this.
A
We
actually
went
back
awarded
bonuses
for
exceptional
content,
submissions
so
for
documentation,
ui
and
videos
that
scored
at
100
of
the
criteria
for
each
of
those
different
categories.
500
phil
as
a
bonus
was
awarded
and
then
whatever
phil
was
remaining
at
the
end
was
again
allocated
back
to
the
participating
teams
proportionally,
and
so
that's
how?
If
you
look
at
the
previous
table,
these
little
cells
in
green
are
actually
the
scores
where
we're
like
wow.
This
is
like
excellent
met.
A
Every
single
ask
criteria:
the
standard
was
really
high,
and
so
that
team
is,
you
know,
deserving
of
an
additional
bonus
yeah.
Once
again,
congratulations
to
all
the
25
teams
that
participated
well,
more
teams
that
participated
25
teams
that
ended
up
being
eligible
for
rewards
like
really
really
well
done.
It
was
awesome.
It
was
a
privilege
to
get
to
score
the
content
that
you
had
created,
and
so
thanks
for
the
time
that
you
put
into
this
some
quick
takeaways.
Some
of
these
were
also
shared
in
slack.
A
We
did
get
a
retrieval
success
rate
published,
but
it
was
not
actually
included
in
the
scoring
for
phase
2.3,
and
this
is
primarily
because
we
didn't
have
enough
of
a
sample
size
for
all
the
participating
teams.
However,
we
have
started
publishing
this
on
the
website,
so
you
can
always
check
it
out.
If
you're
a
project
team,
you
can
go
and
see
what
our
attempts
at
retrieving
your
data
have
looked
like,
and
this
will
certainly
be
a
part
of
the
scoring
for
phase
2.4
and
likely
future
phases
as
well
in
terms
of
documentation.
A
A
Sometimes
team
said
just
go
to
the
network
and
get
it
it's
important
to
cover
both
the
cases
because
clients
that
are
coming
and
wanting
to
get
the
data
from
teams
that
have
onboarded
this
data
on
the
slingshot
could
have
a
preference
for
getting
data
directly
from
miners
on
the
network
or
going
over
to
the
uis
that
you've
built,
but
also
once
they've
retrieved
these
individual
pieces.
Constructing
them
back
into
a
usable
data
set
is
key
and
many
times
that
content
was
missing.
A
A
Our
father
has
also
been
doing
some
awesome
work
with
open
sourcing,
pooling
that
has
made
storing
large
amounts
of
data
specifically
for
slingshot,
more
viable,
and
so
her
presentation
will
include
some
of
that
as
well
dear
you
guys,
so
I
personally
really
enjoyed
seeing
some
really
awesome,
javascript
and
front-end
work,
and
so
congratulations
to
those
of
you
like
that.
That
put
in
the
time
to
do
that,
it
was
awesome.
I
really
enjoyed
it.
A
I'm
sorry,
I'm
rushing
through
some
of
this
as
well.
I
want
to
make
sure
we
get
through
the
rest
of
the
content
so
yeah.
If
you
have
any
questions
about
what
I'm
talking
about,
please
feel
free
to
get
get
to
them
and
chat
and
I'll
definitely
address
them
as
we
reach
at
the
end
of
this
particular
session
on
the
videos
also
fantastic
work
for
those
of
you
that
submitted
videos.
It
was
really
great
great
explanations.
A
Understanding
of
the
data
set,
as
you
know,
showing
the
entire
flow
that
you
went
through
main
flag
here,
similar
to
the
documentation,
like
most
videos,
went
all
the
way
to
showing
the
retrieval
actually
being
done
and
like
a
file
being
obtained
from
the
network,
but
then
actually
like
opening
the
retrieved
content
like
not
just
showing
like,
what's
in
the
dar
file,
but
putting
it
together
and
opening
it
up
in
a
viable
like
application.
So
a
good
example
would
be
for
some
people
that
stored
like
the
linux
iso
database.
A
They
just
show
like
the
the
dar
file
and
then
you'd
pop
it
open
and
you'd,
see
yeah
like
there's
a
bunch
of
images
here.
But
what
do
you
actually
do
with
that?
Because
the
client
is
going
to
want
to
do
something
so
like
mounting
that
somewhere
or
showing
just
you
know
for
30
seconds
what
it
would
look
like
to
extract
and
use
that
to
create
like
a
vm
or
anything
else
would
probably
be
useful,
but
yeah.
Overall,
I
mean
this
was
one
of
the
most
difficult
part
phases
to
participate
in
expectations
were
really
high.
A
Teams
definitely
met
those
expectations.
27
bonuses
were
awarded
across
these
teams,
so
these
bonuses,
as
a
reminder,
were
based
on
exceptional
quality
of
content
submitted
so
yeah.
Congratulations
to
those
teams
that
also
you
know,
met
that
bar
across
multiple
fronts,
and-
and
hopefully
you
know,
you
feel
like
your
work
was
well
reported
and
we
hope
to
continue
seeing
that
high
quality
from
you.
So
what
happens?
Next
slingshot
phase
2.4,
so
phase
2.4
actually
started
as
soon
as
phase
2.3
ended
pretty
much,
and
so
they
already
teams
that
are
onboarding.
A
Data
want
to
call
out
that,
like
today,
congratulations
to
those
that
are
participating,
you
hit
the
milestone
for
15
peppy
bytes,
which
is
25
000
fill
being
distributed
in
rewards
there'll,
be
a
continued
focus
in
phase
2.4
under
capability
submissions,
you
know,
will
include
bill
data
documentation,
deal
ui,
demo,
videos
similar
to
phase
2.3.
We'll
definitely
include
like
the
rsr
number
as
well.
A
One
thing
we're
doing,
though,
is
we're.
Reintroducing
a
limit
of
having
three
teams
use
the
same
data
set.
We
want
to
continue
encouraging
the
diversity,
the
information
being
onboarded
onto
the
network
and,
of
course,
if
you
have
a
data
set,
you
specifically
want
the
on
board
as
part
of
your
participation
slingshot,
and
it's
not
on
the
list
of
data
sets.
Please
feel
free
to
head
over
and
create
an
issue.
So
we
can
have
a
discussion
about
that
so
yeah.
A
If
you're
interested
in
participating,
I
would
say
head
over
to
slingshot.filecoin.io
to
see
the
latest
on
the
competition
and
register
there
to
participate.
If
you
have
any
questions
reach
out
in
the
slingshot
channel
on
falcon
slack
and
I'll,
be
sure-
or
you
know,
somebody
who's
more
knowledgeable
than
me
will
be
sure
to
address
them.
A
So
with
that
thanks
so
much
for
for
being
here,
I'm
going
to
kick
it
over
first
to
laura,
so
laura's
actually
submitted
a
video
that
I'll
be
playing
for
her
presentation
but
she's.
Also
here,
if
you
have
any
questions
for
her,
please
feel
free
to
share
in
the
chat
so
yeah.
I'm
gonna
play
a
quick
video
from
laura
laura
thanks.
So
much
for
making
time
for
sharing
this
with
us.
C
Welcome
to
slingshot
phase
2.3
closing
ceremony
and
congratulations
to
all
the
participators.
We
did
a
truly
extraordinary
job
and
thanks
for
inviting
me
to
talk
about
our
slingshot
project
builder,
so,
firstly,
please.
Let
me
do
an
overview
of
my
team.
My
team,
phil
drop,
has
been
the
fulcrum
ecosystem
from
2019.
C
C
Since
this
reason
we
developed
grabs
plate
a
tool
for
slicing
large
data
science
into
graph
slice,
slices,
fit
form
making
view
in
the
filter
network.
Now
it
an
open
source
project
on
github
and
everyone
can
use
it
or
create
issue
or
pr
to
improve
it.
So
we
so
glad
to
know
a
number
of
screen.
Shorters
use
these
tools
in
slingshot
2.3,
and
we
also
have
received
many
feedbacks
from
community
members.
It
means
a
lot
to
us
and
back
to
this
tool.
C
Well,
why
we
need
it
during
the
slingshot
computation
we
found
when
storing
large
data
size,
we
need
to
slice
it
into
small
pieces
to
fit
for
the
size
of
sector
which
could
generally
be
32,
gb
or
64gb,
and
at
first,
especially
in
slingshot
2.2.
We
made
this
data
into
a
large
tarball
check,
this
powerball
into
small
pieces,
and
then
we
made
storage
deals
with
miners
with
these
pieces
on
the
other
side
of
storage.
This
way
is
quite
efficient
and
allow
us
to
start
hundreds
of
tv
daytime
year
months.
C
C
It
took
advantages
of
ipld
protocol
following
the
unified
format,
data
structure.
It
regards
the
data
size
of
its
subdirectory
as
a
big
graph
and
cut
it
into
small
graphs.
Each
small
graph
will
keep
eight
file
system
structure
as
possible
as
it
used
to
be
so
after
that
we
only
need
to
organize
these
small
graphs
into
a
car
file.
C
Besides
magnifies,
a
csv
will
be
created
to
save
the
mypin
with
graph,
slash
name,
halo,
cid
and
pcid,
and
the
inner
file
structure
will
be
supported
in
the
next
release,
which
is
version
0.4.1
and
another
advantage
of
this
tool.
Is
it
can
perfectly
match
ipfs
like
if
you
build
an
fps
website
as
your
deal
ui
website,
the
internal
file
structure
of
each
database
can
be
shown
on
it,
and
it
is
easy
for
users
to
retrieval
and
download
data
that
you
stored
during
a
slingshot
computation.
C
C
D
A
Awesome,
thank
you.
So
much
for
sharing
laura,
really
appreciate
it.
Congrats
again
for
your
participation
in
this
competition.
Your
team
did
extremely
well
and
we're
really
grateful
for
the
work
that
you're
putting
in
in
improving
the
state
of
the
network,
bringing
data
onto
far
coin
and
creating
awesome
resources
like
go
graphs
but
for
other
participating
teams
as
well.
A
E
Hi
everyone-
this
is
charles
from
first
one
team,
so
it's
very
happy
to
working
with
you
guys
during
strange
swing,
shot
2.3
and
science,
lots
of
teams
already
present
how
to
downloading
and
participating
in
the
screenshot.
Here
we
are
bringing
to
another
interesting
topic
about
how
to
use
the
data.
So
so,
first
of
all
we
are
going
to
introduce
about
ourselves
and
we
participated
the
screenshot
from
1.1.
In
the
beginning
we
are
working
as
our
deep
learning
platform.
Then
we
gradually
going
deeper
and
deeper.
E
We
can
close
with
fragile
network,
so
we
finally
started
the
first
one
project
and
then
just
from
last
month
we
are
happy
to
announce
that
first,
one
now
not
just
a
project,
but
it's
also
a
company.
So
we
built
we
are
going
to
build
this
swan
cloud
computing
company
to
dedicate
using
the
decentralized
age
infrastructure
marketplace
and
provide
the
processing
based
on
fico
network.
E
So
with
this
phase
of
screenshot,
we
are
using
the
data
center
common
code,
which
is
a
non-profit
organization,
provides
per
byte
of
data
connected
from
internet
science.
2008.
An
interesting
story
is
that
zip
provides
their
own
data
format,
which
is
wrx
for
my
files
and
with
different
size,
tags
and
information,
including
that
website.
E
So,
while
we
are
using
those
data
set
and
start
to
analyze
it,
we
have
some
different
challenges
as
a
client,
for
example
the
common
crew
organization.
If
they
want
to
backup
their
data
set
to
the
frequent
network,
they
will
have
some
challenges
like
with
the
fire.
Conceptors
are
32
gigabytes
or
34
gigabytes.
How
can
we
doing
trunk
or
merge
data
effectively?
E
E
So
this
is
usually
the
typical
process
for
clients
update
their
site.
First,
they
need
to
prepare
dataset
either
they
get
it
from
the
internet
downloaded
or
they
have
petabytes
of
hardware
drives
get
ready,
so
they
need
the
trunk
and
the
mergers
files
they
need
to
upload
those
files
to
the
server
or
they
need
to
export
it
to
hdd,
which
will
be
shipped
to
data
centers
like
file
coin
discovery
drives,
does
after
they
have
those
facilities
ready,
they
need
to
find
a
couple
of
miners
and
then
they
need
to
make
deals.
E
This
is
the
whole
cycle
lifecycle
on
the
client
side,
and
this
is
the
way,
if
you
do
it
all
manually,
with
the
original
bluetooth
website
and
the
noodles
terminal.
What
you
need
to
do
you
need
using
to
calculate
how
many
characters,
how
many
lengths
of
the
deal
you
need
to
download
it.
You
need
to
target
the
file
to
combine
with
the
size.
E
Then
you
need
a
computing
cid
and
the
files
piece
id,
and
of
that
you
can
start
doing
the
current
import
of
the
input
you
can
use
in
the
claim
dl
and
using
the
pcs
id
and
instead
of
similar
yields
from
this
place
to
the
miner
and
the
end
you
get
is
a
dr
id,
which
you
can
use
it
for
saving
your
your
database
in
the
first
one
we're
doing
in
another
way
about
the
batch
gears.
We
open
source
our
project
on
the
first
one.com
and
you
can
check
it
out
the
documentation
side.
E
If
you
complicate
it,
they
will
give
you
a
guide
about
how
to
do
this
step
by
step
with
first
one
we
created
a
python
repository
allowed
you
to
sending
deals
together
as
a
batch
file.
So
in
first
one
we
define
the
deals
as
tasks
so
with
one
with
some
small
data
set.
You
can
put
all
the
deals
in
one
task.
For
example,
you
can
have
three
or
30
dls
in
one
task
and
you
give
it
a
name
for
you,
it's
easy
to
remember.
E
You
can
also
give
it
a
tag
about
the
different
data
set
to
help
you
to
fetch
it
in
the
future.
We
provided
the
function
to
download
the
entire
data
set
with
all
the
dr
series
together.
So
when
you
have
those
information
ready,
you
can
use
the
python
tool
cri
to
just
point
to
the
directory,
which
contains
your
carfare
and
you
can
send
all
the
batches
with
one
shot.
E
On
the
other
side,
when
miner
got
those
kind
of
deals
and
instead
of
the
online
or
offline
csv
importer
manually,
you
can
also
download
it
directly
from
first
one
website.
It
can
give
you
the
csv
file,
contains
all
the
information
or
you
can
using
the
api
key
directly
connected
to
the
first
one
website.
E
It
will
provide
the
tools
code
using
just
one
command,
prp
install
first
one
miner.
It
will
allows
you
to
install
all
the
dependencies
if
the
embedded
downloading
to
our
real
tool
can
open
a
monitoring
downloading
from
the
address
which
is
you
put
it?
You
upload
it
from
a
client
on
the
website.
It
will
automatically
download
all
the
deals
and
start
importing
as
you
wish
in
the
configuration
file
you
can
define
when
you
want
to
start
and
the
the
waiting
time
between
each
task.
E
As
a
data
consumer,
we
provide
a
certain
ways
to
reorganize
the
data,
so,
for
example,
with
the
common
core,
we
have
a
website
to
give
you
all
the
information
about
the
deals.
All
the
deals
is
combined
to
the
csv,
as
we
have
shown
it
before
with
each
dataset.
You
will
know
that
which
miner
id
and
contains
which
kind
of
file-
if
you
click
it,
it
will
give
you
the
detailed
information
about
all
the
original
file
which
is
merged
to
one
car
file.
E
E
E
Its
purpose
is
search
all
the
web
pages,
which
contains
youtube
videos
to
see
during
this
time
frame
this
month.
How
many
websites
contains
youtube
videos?
It
will
take
a
couple
of
minutes
for
processing
the
data.
So
in
the
beginning,
you
just
set
up
a
radix
formula
to
for
the
python
to
search,
and
you
get
the
data
which
is
downloaded
from
falcon
network
using
the
commander
tool
to
start
a
sub
process,
which
we
are
doing,
the
compress
the
data
of
your
decreased
data.
E
You
set
the
data
set
to
the
read
mode,
then
we
are
doing
the
matches
to
search
through
all
the
packages
to
see
the
content
which
is
matches
the
youtube
links.
So
the
process
takes
a
little
bit
of
time.
Then
you
will
see
that
we
found
there
was
about
three
4242
youtube
links,
matched
in
this
website
56
000
website.
E
A
Thank
you
so
much
alex
thanks
for
also
being
here
and
presenting,
and
you
know
continuing
to
contribute
to
this
competition
through
all
the
phases
looking
forward
to
seeing
what
you
pull
off
for
the
next
phase
cool.
With
that
I'd
like
to
introduce
our
next
speaker.
Faye,
who
worked
on
storing.
Actually
two
different
data
sets,
but
he'll,
be
speaking
specifically
about
the
a2d2
dataset
feel
free
to
take
over.
F
F
F
F
List
of
data
sets
this
one,
and
if
you
go
to
the
website
in
the
download
section,
there
is
a
list
of
all
the
files
that
is
available
for
download
these
ones
and
it's
very
simple
to
retrieve
them,
because
each
of
the
data
set
has
the
url
on
the
ews.
You
can
just
put
the
prefix
before
each
file
name
and
be
able
to
retrieve
them.
F
F
F
I
also
made
a
web
ui
specif
for
each
data
set
that
I've
been
worked
on.
I
will
demo
the
a2d2
dataset
it's
on
this
little
website.
Here
I
focused
on
the
data
usability
or
easy
to
retrieval.
I
made
a
navigation
on
the
left
here
which
looks
very
similar
to
the
official
download
page.
It
contains
pre
preview
segmentations
you
can
find
them
are
organized
very
similarly.
F
F
F
G
F
A
Awesome,
thank
you
so
much
faye
and
nice
work
working
through
the
technical
glitches
that
looks
like
I
came
back
just
in
time
as
well,
so
we
got
some
of
the
great
work
for
you
as
well.
Search
feature
is
fantastic
website.
It's
great
folks
recommend
you
definitely
check
it
out,
especially
if
you're
interested
in
and
seeing
what
a
great
deal
ui
looks
like
thanks
for
being
here.
Faith
really
appreciate
it.
A
So
with
that
I'd
like
to
go
into
the
last
session
from
a
presenting
team,
which
is
from
free,
rainbow
tables,
is
the
data
set
and
julian
is,
is
the
person
who
was
participating?
Some
of
you
may
have
seen
him
on
slack
at
sonic42
julian.
I
know
you
can
make
it,
but
thanks
so
much
for
submitting
this
video
that
we're
able
to
show
instead
and
I
think,
for
quality.
I
think
I
might
just
play
my
local
copy
of
this
instead.
So,
hopefully
you
can
all
hear
this.
H
H
We
are
trusted
by
protocol
labs
being
selected
as
a
five
plus
notary
receiving
two
dev
grants.
Member
of
the
minority
program
and
frequently
invited
as
presenter
for
falcon
events.
We
are
trusted
by
the
community.
More
than
400
miners
have
deployed
forecaster
for
their
day-to-day
activities
and
by
leading
actors
like
textile
and
flick.
H
H
E
H
H
H
H
Now,
let's
go
through
the
process,
we
applied
to
store
the
rainbow
tables
on
filecoin.
First,
we
downloaded
and
bundled
each
set
of
rainbow
tables
separately.
We
kept
the
same
organization
as
the
original
dataset.
If
you
are
familiar
with
the
rainbow
tables
and
already
have
an
incomplete
set,
you
can
download
the
remaining
pieces
from
filecoin.
H
H
To
retrieve
the
file
simply
go
to
the
free
rainbow
tables.twinter
website
select
the
dataset
on
the
left.
Pane
select
the
file
you
want
to
download,
see
the
deal
id
copy,
paste,
the
lotus
command
line
to
a
terminal
and
wait
for
the
download
to
complete
repeat
the
process
for
all
files
you
want
to
download.
H
H
H
A
Thank
you
very
much
thanks
julian
for
sharing
your
video
first.
I
know
that
you
could
make
it
at
this
time,
but
it
was
great.
I
appreciate
you
stepping
out
and
creating
the
recording
for
us
to
share
excellent
work
as
well,
showing
the
retrieval
all
the
way
through,
as
well
as
the
application
in
which
the
files
can
actually
be
used.
A
So
with
that
we're
into
the
last
part
of
today's
summit,
we've
kept
a
few
minutes
for
q
a
but
as
far
as
I'm
seeing
actually
all
the
questions
so
far
that
have
come
in
chat
were
addressed
in
chat
as
well.
I'll,
probably
just
hold
for
a
minute
to
see.
If
there
are
any
other
questions
in
the
room,
if
you're
here,
please
feel
free
to
raise
your
hand
and
share
a
question.
If
not,
I've
got
one
or
two
last
things
I'd
like
to
chat
about,
and
then
we
can
wrap
up.
A
Added
to
the
site,
so
if
you
head
over
to
slingshot.filecoin.io
so
for
those
of
you
that
aren't
as
familiar
with
the
site,
this
is
sort
of
the
home
base
for
the
competition.
You
can
see
the
latest
statistics.
You
can
see
the
rewards
that
are
being
distributed.
You
can
check
out
the
rules
to
participate
the
prizes
upcoming
events
such
as
this
one,
but
I'd
like
to
introduce
you
to
a
new
tab
that
was
just
added
called
explore,
and
so
this
is
slingshot's
data
explorer.
A
Where
you
know
you
can
check
out
the
data,
that's
being
onboarded
onto
the
network
in
the
form
of
you
know
the
data
sets
that
are
eligible,
but
through
the
various
projects
that
are
staring
at
they're,
storing
that
data
on
the
network,
and
so
you
can
discover
where
you
can
actually
obtain
the
files
for
yourself,
so
we
could
probably
grab
one
of
these.
So
perhaps
we
can
grab
linux
iso.
A
This
one
is,
you
know,
as
you
can
see,
being
stored
in
a
bunch
of
different
geographies
lots
of
different
flags
here
being
covered
and
the
nice
part
about
this
one
is,
you
can
see.
The
file
names
are
familiar.
Debian
different
versions
of
it,
fedora
and
so
based
on
this,
you
can
actually
use
this
as
an
explorer
to
find
the
exact
iso
file
that
you'd
like
to
download,
see
which
cid
it
fits
into
and
figure
out.
A
You
know
where
the
miner
is
that
is
near
to
you
that
you
can
actually
grab
this
content
from
and
run
a
retrieval.
So
this
is
still
just
recently
been
added,
somewhat
work
in
progress
still
as
we
continue
to
flush
this
out.
But
if
you
have
any
feedback,
suggestions,
ideas
and
ways
that
we
can
make
this
useful
as
well
as
just
generally
ensure
that
the
data
that's
being
onboarded
onto
file
coin,
is
being
used
as
charles
was
talking
about
just
making
sure
that
people
that
can
use
it
have
access
to
it
and
can
benefit
from.