►
From YouTube: libp2p weekly sync - March 02, 2020
Description
A
Alright,
hello,
everybody
welcome,
welcome
to
March
any
March.
Second
lid
p2p.
Weekly.
Sync,
please
add
your
updates.
If
you
have
any
to
the
weekly
updates
and
the
crypt
pad,
and
if
you
have
agenda
items,
add
those
and
we
will
take
them
as
they
come
since
we
don't
have
any
agenda
items,
will
jump
into
weekly
updates
and
start
it
off
with
our
Sh.
B
Hey
guys
so
on
my
side,
I
will
I
have
finished
the
work
for
so
the
routing
table
PR
for
conductivity
and
maintaining
a
clean
up
got
merged
after
Stephen
approved
it.
After
that,
I
raise
the
PR
for
the
corresponding
DHT
work
and
some
maintenance
that
you
need
to
do
in
the
DHT.
Steven
is
going
to
take
a
look.
The
PR
for
coordinating,
auto
net
dial
back
responses
with
the
actual
inbound
connections
is
also
up
will
is
going
to
take
a
look
at
that
other
than
that.
B
I
spend
some
time
learning
how
auto
networks
I
reviewed
some
alternate
PRS
that
will
had
raised
and
tries
the
alternatives.
My
focus
now,
which
was
about
going
about
to
write,
is
reviewer
deems
efficient
query.
P
are
basically
disjoint
queries
and
terminating
queries
correctly,
so
I
started
taking
a
look
at
that
I'm,
not
blocking
anything.
C
Everyone
so
last
week,
I
addressed
Stardust
revue,
which
Jacob
did
so.
You
will
need
to
do
another
pass
through
it,
but
we
are
near
the
pier
to
get
merged.
They
also
use
a
flue
got
my
PR
for
the
docks
merged.
So
basically,
in
the
docks
website
we
now
already
have.
The
just
would
be
to
be
tutorial
with
the
newest
sink
API
and
not
broken
one.
Then
I
also
started
creating
some
issues
for
just
BTS,
with
discussed
in
a
diaper
fasting
week.
C
And
finally,
we
also
talked
a
little
bit
about
to
pierced
RV
to
me
and
Jacob
and
I'm,
going
to
think
about
it
on
the
route
of
context
and
start
writing.
Some
stuff
and
I
also
went
to
maybe
sync
with
Youssef
about
it,
since
the
Google
pitch
medium
is
making
some
changes
on
piercer
and
that's
it.
For
me,
the
arsh
yeah.
B
Sorry
guys,
just
one
thing
that
I
forgot
to
say
my
update
is
I,
am
also
come
up
with
a
PR
to
migrate.
The
swarm
notifications
to
the
event
bursts
so
right
now
we
listen
to
disconnect
connect
events
on
the
swarm
right
now.
The
plan
is
to
deprecate
that
it
will
still
work.
We
keep
it
backward-compatible,
but
all
they
connect
disconnect
events
will
now
move
to
the
event
bus.
So
it's
almost
there
step
is
going
to
give
me
one
last
review
and
you
should
be
good
after
that.
D
D
No
in
the
barrels
message
reaction
context,
so
basically
what's
happening
is
that
we
have
effectively
a
special
for
sports.
So
if
a
peer
gets
a
negative
score,
gets
your
move
fruit
from
the
mess-
and
we
don't
add
it.
You
know
that
one
rule
we
need
to
add
your
beers,
okay
and
we
have
a
second
threshold,
which
is
the
gossip
threshold
which
is
slightly
below
zero.
So
this
is
comparable,
which.
D
Which
the
period
becomes
active
with
really
and
we
stopped
even
announcing
positive
or
accepting
goes
from
them,
so
the
PR
is
open.
Your
fridge
review
so
next
up
name
when
publishing
and
the
great
cause
of
tomato
nation
parameters
from
be
adaptive.
Now,
okay
and
then
I
have
to
fill
the
details
of
this.
Far
before
we
read
start
acting
a
five-date.
E
There
we
go
right,
so
we
have
the
discussion
about
removing
the
hench
extra
noise
and
figured
out
that
we
do
want
remove
them.
I
made
a
PR
to
remove
them
from
the
complementation
I'm
gonna
follow
up
today,
because
I
the
noise
library
they
want
to
switch
to
I
finally
made
a
branch
to
use
it
instead
and
it's
slightly
faster,
but
also
much
nicer
and
easier
to
make
faster
still
in
the
future,
because
then
we
can
hook
the
buffer
pool
up
into
it.
E
You
want
to
point
in
the
future
so
I'm
going
to
put
that
up
today,
so
you
can
look
at
it,
but
yeah.
Overall,
the
code
is
much
nicer,
since
we
don't
have
all
this
messing
about
with
nice
pipes
and
I
can
so
so
I'd
like
to
actually
tag
a
release.
Once
we
merge
this
in
forego,
so
you
know
we
can
have
a
number
like
zero,
zero
one
or
something,
and
then
I'm
gonna
also
try
to
pull
my
I.
E
Have
an
old
branch
of
ipfs
that
uses
mice
behind
an
experimental
flag
and
I
could
push
that
up
this
week.
Also,
then,
I
had
also
finished.
The
stack
like
I
ripped
out
a
bunch
of
stuff
in
spec
on
Friday,
but
it's
kind
of
messy
and
there's
some
tangling
references
still
to
mice,
pipes
and
I
finished
that
push
it
up
for
you.
E
I
sort
of
started
just
to
write
out
like
a
much
more
like
you
know,
in-depth,
like
here's
an
example,
handshake
message
and
what
all
the
pieces
are
that's
kind
of
like
in
flight,
just
in
the
background
that
I'm
mentioning
now
so
that
I
won't
forget
that
I
did
it
entirely
and
hopefully
I'll
actually
get
done.
And
then
this
address
Providence
thing
Raul
came
up
with
some
good
stuff
this
morning
and
he
bounced
off
me
and
I.
Think
I'm
going
to
try.
E
Take
that
put
up
like
here's,
how
the
API
would
look,
get
everybody
to
thumbs
up
thumb
it
up
and
then
implement
it
with
your
store.
Try
to
finish
that
work
off
this
week
and
then
probably
you
know
this
week:
I'm
gonna
start
focusing
to
review
visa
stuff
and
get
in
the
Gospels
of
hardening
game,
because
that's
probably
something
I'm
gonna
be
spending
time
on
in
the
next
several
weeks.
That's
what's
up
with
me.
Okay,
that's
better!.
F
E
F
G
Totally
fine,
we
we
had
a
chat
this
morning
and
agreed
that
you
know
don't
think.
Let's
implementation
would
be
the
way
to
look
for
it
here,
but
the
way
that
we
do
it,
that's
what
if
we
decide
to
night
merge
whatever
is
in
progress
now
and
then
Stevie
one.
Do
you
want
a
PR
to
be
submitted
to
ID
FS
as
well
to
incorporate
this
as
an
experiment?
Well,.
G
H
So
have
some
discussions
last
week
with
Steven
about
how
to
handle
people
who
wanna
run
dhts
over
lands
or
VPNs.
They
plan
just
talking
with
to
beat
about
kind
of
just
about
some
of
the
short-term
proposals
for
tht
improvements,
although,
given
that
they're
planning
on
doing
an
RC
for
0.5
very
soon
I,
don't
know
that
really
is
gonna,
be
super
necessary.
H
H
So
I've
put
together,
like
a
number
of
branches,
to
try
and
separately
tackle
each
of
the
new
things
that
we
wanted,
that
we
have
been
proposing
and
working
off
of
and
making
sure
we
actually
get
out
of
it.
What
we,
what
we
want
and
then
sort
of
adding
in
the
optional
functionality,
we
need
to
support
things
like
lens
and
BPMS
as
this
happened.
H
So
we
have
one
branch
which
has
query
logic:
another
which
has
mode
switching
between
client
and
server,
based
on
events
coming
over
the
event
bus
another
that
has
using
events
from
the
event
bus
instead
of
the
notifications
to
add
peers
to
your
routing
table
and
then
will
be
another
one
about.
You
know,
filtering
peers
out
there,
also
a
bunch
of
like
random
other.
You
know
commits
that
seem
to
have
made
their
way
into
the
branch
that
I'm
gonna
be
parsing
through
and
in
putting
up
there.
H
So
that
means
that
if
you
are
one
of
the
people
who
made
one
of
these
commits
in
the
last
six
months,
you'll
be
getting
tagged
for
review
to
make
sure
that
what
was
relevant
from
when
you
did
this
originally
is
still
there.
And
if
you
are
building
things
on
top
of
this,
which
I
think
is
mostly
harsh.
H
The
stablize
putting
things
on
the
stabilized
branch
is
not
going
to
be
very
helpful
to
you,
because
then
you're
just
gonna
have
to
rebase
this,
and
do
this
all
again
later.
So
I
would,
I
would
watch
out
for
those
branches
and
see
if
they
look
like
they
are
missing
things
to
you,
because
you'll
have
had
more
time
to
look
into
the
code
and
you'll
be
like
wait
that
wasn't
there.
When
I
looked
at
it
and
then
you'll
know
that
I
missed
a
commit
somewhere.
B
B
B
A
F
A
I
G
0.3
and
taking
off
this
plane
officially
tomorrow,
prob,
so
that
that's
one
other
way,
I
waked
on,
as
Yusuf
said
on,
address
permanence.
On
modeling
this
we
have
like
a
good
way
forward
to
models,
Benina
that,
if
I'm
gonna
be
handled
by
the
pier
store,
we
probably
gonna
need
to
add
some
some
options
and
functional
options
to
some
methods
in
the
pier
store
and
a
few
other
things,
so
I'm
gonna
be
submitting
a
work-in-progress
PR
just
so
that
you
can
build
on
top
of
it.
This
particular
data
type
I.
G
G
Gonna
also
also
did
some
work
in
the
in
the
gossips
of
securing
gossips
about
task
force
to
prepare
things,
and
they
the
way
for
for
the
work,
that's
happening,
and
it
turns
out
that
github
sort
of
happen.
His
hogs
the
issue
with
the
problem
with
issue
migration,
so
I
think
we
might
be
at
a
good
spot
now
to
try.
You
know
the
the
whole
thing
again
they
so
the
issues
that
were
the
issues
that
were
bored
are
still
for
it.
G
They
haven't
fixed
that,
but
apparently
they
claim
that
if
you
represent,
if
you
do
the
same,
if
you
do
the
same
step
by
step,
it
would
lead
to
the
same
result.
The
problem
is
that
I
can't
access
the
original
issue,
so
I
remember
how
I
model
those
and
to
reproduce
that
so
anyway,
we'll
see
how
we
got
a
confirmation,
because
we
definitely
don't
want
to
end
up
working
the
entire
issues.
Often
it
b2b,
if
we
if
we
go
for,
and
that
would
be,
though,
basically
throw
us
off-track
for
a
long
time.
G
So
this
week
is
gonna,
be
a
court
intensive
week
for
me,
I've
been
I've,
been
knowing
a
lot
of
people,
a
lot
of
reviews
and
code
and
collaboration
and
a
lot
of
things
in
general,
so
I'm
gonna
try
to
catch
up
with
all
of
that
this
week.
This
means
that
I'm
gonna
be
doing
a
hardcore
sprint
on
test
ground,
with
a
DHT
test
plan
and
with
debugging
all
the
way
and
enjoying
all
the
way
and
to
understand
where
the
bottlenecks
are
to
understand
what
we
need
to
polish.
G
What
we
need
to
optimize,
I'm
gonna,
be
developing
on
the
address
provenance
changes
and
periodic
records
with
with
with
Yousef.
We
really
want
to
learn
this
this
week.
The
only
block
and
change
blocking
thing
was
there
was
the
peer
store
address?
Probably
this
thing
so
with
that
out
of
the
way
we
hope
to
land
us
so
that
Yousef
can
also
focus
on
gossip
self
would
be
so
next
we
are
starting
next
week,
I
see
you
bobbing
your
head,
Yousef
I,
don't
know
what
that
means.
G
B
G
You
had
this
new
feature
in
the
ps4,
because
for
now
the
dialer
would
just
be
will
will
be
making
this
backwards
compatible.
So
the
dialer
would
just
be
dialing
everything.
That's
that's
available
in
the
Pierre
story,
but
if
we
want
to
find
you
in
tandem,
we're
definitely
gonna
have
to
like
yeah
yeah.
G
C
G
G
What
it's
secure,
a
dining
for
for
0
X
gonna,
be
getting
up
to
speed
with
everything
that
has
happened
in
content
routing
and
the
implementation,
which
is
connect.
The
speed
of
lightning
so
I
want
to
I
want
to
get
up
to
speed
with
that
I'm
joining
the
eath
to
networking
call
as
well
on
Wednesday
and
gotta
Bieber
being
Europe
goes
quality
transport
PR!
G
F
F
D
G
D
D
Yeah,
yes,
so
what
is
already
there
and
squirty
starting
to
review
now?
Because
this
is
how
it
the
Peace
Corps
plugs
into
the
concepts
of
operations.
Yeah,
okay,
yeah,
the
details
of
the
brisk
are
not
so
Dean.
It's
gonna,
be
we're
going
to
do
at
first
I
thought
you,
the
idea
of
doing
a
period
topic
score,
but
I
realize
that
it
was
using
the
global
score
and
some
occasions
may
be
the
topic
weights
within
this
car,
because,
if
you're
misbehaving
in
one
topic,
you
should
be
penalized.
D
G
D
F
F
D
G
G
Be
I'm
gonna
be
reaching
out
to
you
very
likely
if
you
have
any
changes
that
are
implied
that
you
haven't
landed,
that
you
haven't
committed
a
pressure
upstream.
Yet
then
I
would
like
to
like
set
a
good
Check
Point
so
that
I
take
your
your
current
plan
and
like
start
debugging
on
it,
and
make
sure
that
you're
not
making.
You
know
changes
in
the
mean
time.
So,
just
like
a
hand
off
point
would
be
great
and.
F
G
G
Once
we
also,
there
are
a
few
changes
that
are
I,
think
I've
been
in
0.5
that
are
gonna
affect.
You
know
the
whole
process,
the
way
that
we
process
the
routing
table
anyway,
because
now
we're
gonna
have
like,
potentially
you
just
connected
nodes
or
whatever
in
the
routing
table.
I,
don't
know
if
we
want
to
want
to
process
those
so
Stephen
and
I
were
actually
in
in
New
York.
G
We
were
discussing
maybe
a
different
way
to
get
to
acquire
similar
information,
so
that
really,
if
you
take
a
step
back
and
look
at
what
we're
trying
to
accomplish
here,
is
bootstrap
and
see
the
routing
table
at
the
initial
stage
when
we
start
the
system
and
also
heal
from
a
from
a
flawed
situation
right.
So
when
there's
a
disaster
or
whatever
reason
that
we
are
been
partition
or
disconnected
from
the
network,
we
want
to
be
able
to
heal
right.
So
if
we
like.
G
G
H
G
So
we
we
took
basically-
and
there
are
two
reasons
why
you
really
want
to
have
this
construct
and
the
DHD
you,
which
is
one
to
start
from
a
blank
routing
table
when
you
start
the
system
right
and
you
wanna,
you
want
to
be
able
as
well
to
recover
from
a
failed
state
of
the
system
right.
So
these
are
the
two
reasons.
G
While
you
want
to
use
some
information
from
a
previous
state
that
you
deemed
healthy
of
the
routing
table
to
recover
from
that
with
some
seedings
tragedy
right,
so
another
construct
to
make
another
approach
instead
of
like
periodically
saving
snapshots
and
then
recovering
from
them
is
to
have
kind
of
like
a
like
a
more
continuous
nursing
model,
no
siree
model
right.
So
it's
like
you
have
a
component
in
the
system.
G
That's
watching
the
connection,
just
connection
events,
the
routing
table
addition,
removal,
evictions
events
and
is
recording
this
information
so
that
when
something
in
the
DHT
needs
a
number
of
peers
to
heal
from
a
failed
state,
they
can
just
request
and
say:
hey
I
want
five
years
on
110
peers
have
112
years,
and
this
construct
is
gonna
return.
You
know
that
many
peers,
according
to
maybe
some
heuristic,
so
it's
kind
of
like
a
different
way
of
thinking
about
it,
that
is
less
overfitting
to
what
we
actually
need
and
it's
maybe
a
bit
more
forward-looking.
H
Especially
like
it
started
up
zero
five,
maybe
even
four
zero,
six
making
sure
that
some
of
this
maybe
is
like
opt-in,
because
it's
it
is
nice
that
it
is
nice,
if
not
somewhat
unfortunate
and
centralized
that
if
really
bad
things
are
happening
to
the
network,
having
control
over
the
bootstrappers
allows
for
healing
and
some
circumstances,
and
so,
if
you're
doing
major
protocol
changes-
and
you
know
before
we
have
like
tons
of
public
evidence
in
addition
to
the
test
ground
evidence
of
stability,
it's
like
nice
to
still
have
that.
I.
F
G
So
really
you
don't
want
to
like
if
you've
been
poisoned
and
you've
been
attacked
and
you
have
a
poisoned
routing
table
you,
for
example,
don't
want
to
be
saving
snapshots
at
that
table
because
then
you're
gonna
be
attempted,
so
currently
the
PRA
that's
up
would
be
having
this
problem
right.
So
that's
why
this
forced
us
into
like
rethinking
a
little
bit
more.
Would
like
all
the
information
that
we
have
all.
E
G
F
F
Really,
the
important
part
is
that,
like
box
asks
pierced
our
changes
and
trying
to
get
the
noise
changes
murder,
so
we
can
forgive
as
a
whole.
We
actually
think
at
this
point,
maybe
like
deeper
ties,
enjoy
stages
a
little
bit
for
like
a
week
or
two
might
be
a
good
approach,
because
I
don't
think
like
we're.
Gonna
get
them
to
like
the
good
place,
want
them
to
be
to
cut
a
release
like
in
the
next
couple
of
weeks
in
terms
of
the
piercer
changes
that
sort
of
talking
about
like
at
the
moment.
G
F
It's
like
my
concern
is
that
I
don't
want
everything
also
to
block
on
this
at
the
moment,
like
everything
is
kind
of
blocked
on
these
piercer
changes,
because
it's
all
kind
of
funneling
through
like
the
core
API,
and
we
can't
really
release
a
new
comp
new
core
because
the
teams
have
made
there.
So
now
we
can't
we
have
some
events.
F
F
The
sinecure
store
stuff
also
has
like
these
events
around
like
generating
new
sign
records
and
that
kind
of
stuff
yeah.
We
just
we're
also
getting
events
for
hey.
Your
addresses
have
changed,
but
you're
planning
on
using
yeah
inside
of
automats
like
it
would
be
nice
if
we
could,
just
like
happens,
anyways
to
say,
okay,
we're
not
actually
storing
the
sign
records
we
can
still
generate
them.
Make
them
speak
the
new
protocols
if
we
need
to
that
kind
of
stuff
yeah.
So
if
we
could
somehow
merge
that
stuff
without
merging
the
rest
of
it,
we
kind.
B
I
H
A
H
F
A
F
All
like
this
thing,
we
can
still
start
merging
all
three
change
together,
like
testing
them
all
as
a
unit
so
like.
If
this
thing
magically
shows
up
and
like
we
actually
interest
actually
finish,
it
think
great,
but
it
doesn't
feel
like
it'll
cause
any
issues.
Then
we
could
try
sneak
in
yeah
but
like.